Edit: This most resembles the old Administrator access-level; able to add/edit/delete accounts, properties, profiles, filters, goals, etc. However they are unable to add/edit/delete user.

Manage Users: This is a lighter administrator setting and can only add/delete accounts and users, as well as assign user permissions. These users cannot edit or view reports.

View: Just like the ‘Users’ from the previous settings, this setting only allows users to view report data.

image from http://analytics.blogspot.com/

These three user settings can set in any combination together to give the user the most access they need, while also restricting the user where necessary.

One important consideration with the permissions is that they are inherited at the parent level (account > property > profile). Therefore, when permissions are set for a user at the account level, those permissions carry down through the property and profile level for that user. This is part of the automatic settings, but as mentioned above, they can be overridden for a customized access-level.

During the migration process, all users that are currently set with admin access will be switched to full (manage users, edit, and view) access while the old user access will be changed to View only with access to just the profiles they previously were allowed to see.

With these changes, companies using Google Analytics will be able to have greater control over the users accessing their data through the use of the custom combination settings, meeting the needs for the company and the users alike. The ability to restrict specific user permissions at the account, property, or profile level combined with the new change history log will provide the Analytics account owners the assurance that their data is safe and users will be able to see only the data they need to.

This new feature has been a part of AdWords for a while; it is new and exciting within Google Analytics. Change History will be available to all account administrators over the next few weeks, allowing them to see changes made in their accounts for the past 180 days.

Image from analytics.blogspot.com 

This element in Google Analytics will be housed within the Account Level of the Admin tab. In here users will be able to see what the change was, who made it and when it happened. Change History will track updates to users, accounts, properties, goals and filters. Knowing this information will keep web analysts like me in the loop when something changes on the site, for better or worse.

There have been many situations where I wish that this feature was available in the past. One of most common uses would be verification that a user has been created or removed in the account. This new feature shows changes to user statuses and being able to send a screenshot to a manager for verification would be very beneficial. Another useful element with the Change History is the timestamp feature, now we will be able to know when we changed the name and/or the definition of goal. Throughout the life of a website, goals become outdated and need to be replaced. This feature will allow administrators to look back during that 180 day window and find out if any of the goals changed and can log the changes in the annotations section of Analytics or in an external document.

What exactly are enhanced campaigns?

Google highlighted three major changes rolled into Enhanced Campaigns:

1. Simplified campaign & budget management with the ability to change and manage bids across device, location, time of day and more within a single campaign. Previously, Google had recommended splitting these targets into separate campaigns to allow for more targeted bidding, since bids could not be changed on individual keywords based on targets.

2. Smarter and more customized ad delivery, allowing advertisers to show the best ad text and ad extensions based on where, when and how a person is searching, without having to segment campaigns for every possible combination of devices, locations and time of day.

But these changes weren’t the only additions. Google seemed to skim over the biggest (or at least the most anticipated) change that was rolled into enhanced campaigns – extension updates. With Enhanced Campaigns, extensions will be more flexible and more targeted. And, after years of waiting, advertisers can now report on individual sitelinks and create extensions at the ad group level (as opposed to at the campaign level).

What does that all mean?

Pros

• Generally easier campaign management across devices and locations
• Deeper and more specific targeting options can be explored due to easier management and optimization
• Huge improvements to sitelinks and sitelink reporting
• Mobile traffic is emphasized
• Ads can be automatically optimized

Cons

• Loss of control and transparency over campaigns and targeting
• Tablets are removed as a targeted device and lumped in with desktops
• No more mobile-only campaigns
• Could have an immediate and sizable negative impact on KPIs for sites that are not mobile-friendly
• Mobile CPCs will potentially increase significantly due to increased competition
• Merging campaigns, updating settings and reworking optimizations when campaigns transition to enhanced campaigns could be tedious

Why should I care & what can I expect?

Greater Targeting Options & Account Efficiencies

Smaller businesses and advertising budgets will benefit from the ability to target specifically and incrementally within the same budget. Before, creating hundreds of campaigns to target different DMAs and devices would have split the budget too finely to see results. With Enhanced Campaigns targeting can be achieved using the same budget and campaign, by simply modifying CPCs.

Larger advertisers will see increased efficiencies that allow even further targeting and ad testing that was previously not possible (either due to maxing out on number of allowed campaigns or due to overall manageability and control). However, these advertisers will be forced to give up some of the control that they are used to. In addition, while all advertisers should see efficiencies from a management perspective, this may not translate into being more efficient from an ROAS perspectives due to decreased budget management options and increased CPCs.

Increased Focus on Mobile

Sites that are not mobile optimized will be the most negatively impacted by Enhanced Campaigns, since mobile campaigns will no longer be budgeted separately. Assuming advertisers without mobile-friendly sites are still advertising on mobile, but at a reduced budget, mobile advertisements may use budget that was originally allocated only to desktop campaigns, increasing mobile as a percentage of paid search traffic. If your site does not offer a good experience for your mobile users, these clicks could be wasted.

With the migration to Enhanced Campaigns, it is obvious that Google is forcing advertisers into the “future” – which focuses on mobile. Google has forecasted that they believe mobile searches will surpass desktop this year, so expect them to do even more to push mobile and to be a part of this change. Expect a mobile site to be a necessity in the near future.

More Ad Extensions 

All advertisers will see improvements in ad extension reporting that will allow more granular optimization. Sitelinks will be implemented and optimized at the ad group level, which should improve ad relevance and increased CTR. We expect that in addition to sitelinks, there will be an increased use of extensions in general. With the ability to have ad extensions auto-optimized based on ad, context and search, more advertisers will opt in. We also expect for Google to release more extensions in the coming months.

At Proximity, we have a specific definition for each of these terms. In this blog post we hope to clarify the difference between these two terms and show how both play a unique role in digital measurement and analytics.

What are ‘metrics’?

Put simply, the term ‘metric’ refers to any number that attempts to quantify performance. Pageviews, unique visitors, bounce rate, conversion rate, revenue, and average order value are all examples of digital metrics. Alone, one metric is relatively meaningless; it is only when important metrics are given context and  compared to benchmarks and goals that we can gain an accurate and full view of performance and progress. But for any given media campaign, website or promotion there are dozens (if not hundreds) of potential metrics. How do we know which metrics are important? Enter key performance indicators (KPIs).

What are ‘KPIs’?

KPIs are specific metrics that quantify performance relative to business and/or digital objectives. KPIs should be carefully chosen from all available metrics based on a variety of criteria:

1. Do these metrics accurately portray the performance of a specific objective?

2. Do you have access to the metrics that you have chosen for your KPIs?

3. Do the KPIs lead to action?

Ideally, each of your objectives should have at least one corresponding KPI.

For example, imagine that you are a marketing manager at an e-commerce company and one of your objectives is to increase customer loyalty. To accurately measure progress towards this objective, you will want to choose metrics that correspond with customer loyalty.

Below is a list of potential metrics you could select as your KPIs.

• Purchase frequency
• Average order value
• Purchase conversion rate (purchases/visits)
• Return visits
• Pages per visit
• Bounce rate

However, only the top 3 in the list (purchase frequency per year, average order value, and purchase conversion rate) indicate the level of customer loyalty, so you would want to select those metrics as your KPIs. Another objective will have a different set of KPIs.

In a past blog post, we covered the role of metrics and KPIs in creating an actionable  measurement strategy.

Ultimately, choosing the right KPIs is important to accurately evaluating performance. In turn, a strong understanding of performance helps drive meaningful actions and optimizations that bring you closer to your goals.

Users can search across four main categories, using a set of intuitive verbs (“lives,” “like,” “work,” etc.), nouns (“San Francisco,” “Indian,” “restaurants,” “friends” etc.), prepositions (“before,” “with,” “in”) and pronouns (“who,” that,” etc.). “Friends of friends who live in San Francisco and like Indian restaurants,” for example. Or “Friends who have been to Ireland,” or “Photos of friends before 1990.”
- John Battelle

What about traditional search engines?

For Google, this new service represents the organization and distribution of data that Google does not have access to. Sure Google might know your email address, that you like food blogs and the types of things you've searched for lately. However (unless you've completely filled out your Google+ profile) they don't know where you live, the name of your company, specific restaurants you like, how many pictures you are in with specific groups of friends, etc. That being said, Graph Search is not (in its current state) direct competition for Google. There is plenty of information outside of Facebook that people are looking for.

To address this, Facebook turned to its long-time search engine partner – Bing. Graph Search will pull in Bing web results if

it can't find good results within Facebook. Obviously, this is good news for Bing, but it remains to be seen if this will increase people's proclivity towards using Bing when they're not on Facebook. I personally doubt seeing Bing results on Graph Search will change that behavior.

What about traditional SEO?

The two things that fuel Graph Search have been important aspects of a good SEO strategy for years – social media activity and structured data.

Social Media Activity – Likes, Followers, Fans, etc.

For years both Bing and Google have been increasingly incorporating social activity into their search results page using Facebook and Google+ respectively. The SEO benefit is one of the reasons we include social sharing buttons where appropriate, create engaging content that our audience will want to share, and connect with fans and followers on social networks. The things you and your friends seem to like have a better chance of appearing in your search results. This still matters; it might just matter a bit more now.

Structured Data

Manually labeling individual words, phrases and images using microdata is not new. It's an SEO tactic that, in all likelihood, will continue to be important for some ti

me. For traditional search engines, structured data helps Google understand how many calories are in a recipe or how many products are in stock. For Graph Search, structured data helps Facebook understand your relationship status, exact brands that you like, and where you geographically tag yourself (i.e., where you 'check-in'). The big difference, is that Facebook has easy access to structured data, since almost everything its users do from filling out their profile, tagging photos, liking brands, etc. is automatically categorized and organized whereas Google has to work for it – constantly crawling pages and requiring additional programming of sites' developers.

(Read more about microdata and SEO)

Advertising Opportunities

The ability to reach Facebook users based not only on their profile data (as you can do now) but also Graph Search behavior will open the door to many hyper-targeted advertising possibilities. One of the reasons that Google has been so effective at matching advertisers with an audience is its ability to predict what that audience is going to do in the future. For example, searching for recipes indicates that you might be planning a grocery run and searching for headache medicine might indicate that you need a pain reliever in the near future. While this seems fairly obvious, it is this insight into future behavior that has made Google a lot of money.

Facebook, on the other hand, has relied on past behavior for its advertising – people who check-in at their favorite lunch spot probably won't be receptive to ads for nearby restaurant, posting photos from vacation doesn't indicate the need for hotel deals and liking a brand provides more post-purchase/retention opportunities than those that increase new customers/awareness. Graph Search could change that by giving Facebook the future-focused behavioral data that traditional search engines have thrived on.

The more things change the more things stay the same

Clichés aside, many of the same SEO/Social Media priorities remain the same for brands and publishers:

• Encourage social activity by creating great, sharable content
• Use Microdata to create structured data on your own site. This will help with traditional organic search as well as when web results are included in Graph Search.
• Create and maintain social media properties that attract and engage consumers
• It might be tempting to engage in tactics that artificially inflate likes/fans through bribery, procurement or otherwise 'spammy' tactics in order to appear more often in Graph Search results. This is a short-term strategy and likely to back-fire as the Graph Search algorythm becomes more sophisticated (similar to Google's Penguin and Panda)

As Graph Search is rolled out it will be interesting to see how likes, comments, posts, profile data, Spotify playlists, etc. play a part in the results. And it gives traditional SEOs another algorithm to monitor, analyze and optimize for, which (to be clear) is exciting…at least for us.

More importantly, there is increased interest in proving the value and impact of social media. The query “facebook best practices” returns 370,000,000 results on Google and “value of a Facebook fan” returns 161,000,000. In short, social media is often a requirement to compete in the digital space and brands want to make sure they are measuring up.

Yet one looming barrier remains: gathering and manipulating social media data can be cumbersome, time-consuming, and costly if you use a social media analytics vendor. But it doesn’t need to be that way.

The aim of this post is to show how you can discover key learnings about the following elements and apply them to your own Facebook marketing/editorial calendars.

• Post type (link vs. status update vs. photo)
• Post length
• Calls to action
• Time of the day
• Day of the week

For example, you may publish posts that are meant to engage your fans with a question or fill in the blank. This blog post gives you the tools to understand when your fans are most likely to engage.

Alternatively, you may be an online retailer and publish posts with links to your deals or weekly specials. This blog post gives you the tools to understand when your fans are most likely to be reached.

Below are the steps for analyzing the aforementioned post elements.

1. Download & Clean Up

First, access your Page’s Facebook Insights and download the Page Level data for the time period in which you are interested. The more data you are able to download, the better.

Facebook Insights Interface

Facebook Insights Post Level Export

Time & Day

Within the Excel export, add a column titled “Day”. Next, fill in the day that corresponds to each date. The formula for this is: =TEXT([cell that contains the date],”dddd”). Here I am using =TEXT(H3,”dddd”). Drag this down for each row in the spreadsheet.

Separate date and time into different columns (column header is “Posted”):

• Find and replace “ PM” with “PM” – and repeat for AM.
• Select the column titled “Posted”
• Pull up the Text to Columns tool, select space as the delimiter and choose a destination column that does not contain data (such as Column E)
• Ensure the columns are split correctly

Label the columns accordingly. Add an additional column titled “Rounded Time” and fill out based on the time closest to the hour mark.

Post Length

Add another column to the right of the “Post Message” column and title it “Character Count.” Insert the length formula, =LEN(), a into the first cell of this column and select the corresponding cell in the “Post Message” column. Drag the formula down. Add one more column to house the post length. To start, calculate if the post has greater than or less than 150 characters. Here we calculate this using the =IF(COUNTIF()) formula. Past studies have shown that fewer than 150 characters is the ideal length, so this column will help us identify the benchmark and performance based on it.

Post Objective

If each post has a specific objective (e.g. engage or drive sales, etc.), add another column titled “Objective.”

The first several columns of your data sheet should now look like this:

2. Build the Pivot Table

Let’s start by analyzing day of the week. Create a new tab. Open the Pivot Table builder and add in the Row Label (here we are using day of the week) and Values (the metrics that you want to analyze). In this case, we are using the following metrics:

• Total Impressions
• Post stories (likes, comments, shares, etc.)
• Post consumptions (photo views, link clicks, etc. plus stories)
• Negative feedback (post hides, report as spam, etc.)

Be sure that the Values listed above are set to Sum; by default, these are set to Count. However, note that Count can be used to tell you the number of posts by pulling in any column name (for example Count of Lifetime Impressions).

The Pivot Table will look like this:

These steps can be repeated for any element in which you are interested: day, time of day, objective, post length range, etc. We recommend creating a new tab for each element, which will make it easier to organize and find data later on.

3. Calculate metrics

Now that the data is aggregated, we can proceed with calculating metrics and KPIs for each day of the week. In this case I would like to calculate the following metrics:

• Average engagement per post (lifetime stories/lifetime impressions). In other words, the number of interactions for each time the post was seen. High engagement rates indicate strong engagement from your fans – the higher the better.
• Average consumption rate (lifetime consumptions/lifetime impressions). This tells us the total number of clicks on your post, including the number of times a photo was viewed or a link was clicked for each time the post was seen. Again, the higher the better.
• Average negative feedback rate (lifetime negative feedback/lifetime impressions).

  While likes and comments are generally indicative of good content, negative feedback shows us how many times the post was hidden and/or reported as spam out of the number of times it was seen. If you are posting repetitive content or the tone of your posts is not consistent with your brand, you will probably experience a high negative feedback. Generally, the higher your engagement rate, the lower your negative feedback rate.

• Average impressions per post (Sum of Lifetime Impressions/Count of Lifetime Impressions) tells us how many times posts were seen.
• Average reach per post (Sum of Lifetime Reach/Count of Lifetime Impressions) tells us the average unique number of people reached per post.

The image below shows the top performing days for the aforementioned metrics.

4. Draw out key findings

OR

2. Your fans are more likely to engage on these days

Similarly, if you publish promoted (paid) posts on particular days, those days will have higher reach and impressions. If you post photos or links more often on particular days, those days will likely have a higher consumption rate.

Start out with data observations and after taking all information into account, pull out key findings. For example, the community manager on my brand tries to post during a variety of days and times. She also evenly publishes photos, questions, etc. So, I can safely assume that the data is not skewed.

When doing the analysis for post type (photo, status update, link, etc.) I discovered that Photo posts have 3x the reach of Question/Poll posts. But, Question/Poll posts see 2x the consumption rate of Photo posts. So, if I want to maximize reach to share info about deals, Photo posts are my best bet. But if I want to engage my fans, a Question/Poll post would be more appropriate.

Facebook Insights and social media analytics in general can be confusing. And you don’t need to blindly follow best practices to be successful in social media. Best practices may give you a good place to start, but using Pivot Tables to isolate attributes of your best performing posts can quickly pay off.

At Proximity, we provide analysis and recommendations for day/time, post length, content, and several other ongoing optimization techniques within a reporting cadence that makes the most sense for each client. As a result, the posts are well-optimized to deliver on our clients’ social objectives and (ultimately) their business goals.

thermore, the idea of “adaptive web design” is starting to take hold, which layers on consideration for device capabilities when building responsive designs. But does Google even care and will this help your

SEO strategy? The short answer: yes.

Responsive Design: The New Benchmark

Aside from reaping gobs of cash from search & media advertisements, Google’s stated goal is to catalog the web in a convenient way for web searchers. Along with this is a focus on user experience and serving up the best possible results. Web queries on mobile or tablet devices often have a very different context than desktop searches. A certain amount of local intent is almost inherent in mobile searches, as consumers are physically on the move and perhaps in a closer position to purchase a product, find a restaurant, check on travel plans, etc. But this is not limited to location-based queries, as searchers do not often self-identify intent in their queries.

In December 2011, Google announced the new Googlebot-Mobile crawler that would function similar to the traditional Googlebot crawler (discovering web pages, indexing & ranking them) but with specific attention to the mobile experience. From the official announcement: “The content crawled by smartphone Googlebot-Mobile will be used primarily to improve the user experience on mobile search. For example, the new crawler may discover content specifically optimized to be browsed on smartphones as well as smartphone-specific redirects.” The fact that Google would create a dedicated crawler indicates their seriousness on the future of mobile and its place in the index.

Making Google’s Job Easier

So we know that Google cares very much about user experience at a device-specific level. If webmasters were to provide content uniquely suited to mobile/tablet/desktop, that would make Google’s job much easier. Google even released their official opinion in June 2012 with a very direct point that responsive “is Google’s recommended configuration.” So if your site requires a horizontal scroll bar on a mobile device or otherwise loads poorly on non-desktop browsers, rank decreases on these devices might occur, with a potential dotted-line to decreased desktop ranking under the penalty of general poor user experience.

Adaptive Web Design Considerations

The concept of “adaptive” web design is relatively new to the scene. This approach not only changes design based on device screen size but also considers the capabilities of that devices, such as touchscreen motions, geo-location or other native features. This might involve scrolling through pages with a finger-swipe or pinching to close a window; actions that currently responsive design does not take into consideration.

The Future of Mobile SEO

While the initial rank impacts affected searches with local intent, Google has indicated the future should include device consideration, and we should expect rankings on desktop/mobile/tablet to begin deviating based on device UX. Catching the early train on adaptive web design could offer additional benefits as this becomes part of Google’s consideration set as well.

Progressive design for the sake of progress alone can be a risky strategy, though with Google publically endorsing responsive design, and a large number of brands adopting as a new baseline of the web experience, it’s a safe bet that responsive (and adaptive) will continue to impact SEO well into the future.

Site quality continues to be a top priority, with updates to freshness, page quality and the Panda algorithm. Additionally, there is continued emphasis on improving local search results and site links. Other notable changes include multiple updates to the autocomplete search function, changing the way Google generates search predictions. Interestingly, there has been a notable absence in Social updates since May.

Included in this document is a graph of the various algorithm updates (categorized by type) from the past 6 months, as well as key highlights from the algorithm updates.

Four edits to page quality and the Panda algorithm

• Description: Three updates related to providing users with content from quality, trustworthy sites. Additionally, Google refreshed data for the Panda algorithm.
• Impact: Google continues to emphasize high-quality sites with original content as the most relevant results to user queries.

Multiple updates to the autocomplete function

• Description: There were several updates to the autocomplete search function, all related to improving how autocomplete predictions are generated.
• Impact: Google will be better able to predict users search queries, and ultimately provide them with more relevant content even quicker than before.

Two updates to “freshness”

• Description: Updates were made to display most recent content when multiple items from the same domain are relevant to users. An additional change improves the way outdated content is handled.
• Impact: It’s becoming increasingly important to provide updated, “fresh” content to users. Outdated, “stale” content will likely result in lower search engine rankings.

Multiple updates to local search results

• Description: Google made multiple improvements to display more relevant local search results, in addition to improving local search features (e.g. local time search).
• Impact: Optimizing for local search is an integral component of SEO as local search queries continue to grow and Google continues to improve its algorithm for local results.

In April 2012, Branded3, a digital marketing agency based in the UK conducted an in-depth study (claimed to be the largest of its kind) to test whether the amount of tweets that include a specific URL affected the page’s ranking on Google SERPs. The agency’s research showed that URLs did, indeed, receive a significant boost when shared on Twitter, but the results were more detailed than that.

It was concluded that improvements in ranking leveled off after roughly 50 tweets and were then virtually nonexistent until around 1,000 tweets. From the 1,000-tweet mark, incremental gains were minimal until approximately 5,000 tweets when the average ranking of URLs increased substantially. Perhaps most promising is that an SEO benchmark was seemingly identified when researchers noticed that URLs receiving over 7,500 mentions on Twitter almost always rank within the first 5 Google results.

For context, the search terms that were considered for this study were the first four words included along with the URL in the tweets. The example given by Branded3 was a tweet “We want Justin Bieber in Salvador, Bahia, Brazil!” was shortened to a four-word phrase (“We want Justin Bieber”) when conducting the experiment in Google searches.

Correlation vs. Causation

While these figures may seem groundbreaking, Branded3 acknowledges explicitly that other variables may have played a role in the findings of this study:

• URLs receiving many tweets are more likely to have an already respectable inbound link profile as they have a better chance of being featured on reputable news sites and blogs.

• The way in which Branded3 created enough tweets to measure was by monitoring a Twitter petition site (Twitition.com), and many of these high-volume Twititions may have been specialized or niche enough to have limited competition, thus rising through Google’s rankings more easily.

As is often the case in SEO, however, the direct impact of implementing a single tactic or best practice is seldom measurable to an exact degree, but a study of this magnitude (8,528 Twititions analyzed), proves that a correlation does exist between a URL’s Google ranking and the number of tweets that URL is included in. And at the very least, the results of this study fall in line with a best practice that we have been preaching for years—establishing an active social media presence for link building purposes and integrating social media sharing buttons into your site for UX purposes.

Implications for Brands

All brands understand the importance social media plays in their marketing mix, but few have plans to actively encourage social sharing of links to the pages they generally promote passively. While nothing is guaranteed in SEO, social engagement tactics like “retweet to enter” contests, or partnerships with major influencers on Twitter could prove to be a serious boon to a page’s organic search rankings.

• 34% of marketers reported that analytics were not integrated at all with their business plan.
• Less than half of analytics data collected is actually useful for decision making, according to a majority of surveyed marketers.
• Only 11% of marketers’ customer-related decisions are driven by data, according to a CEB study. The same study claims that a majority of marketers still rely too heavily on intuition.

Clearly, creating and executing an actionable web analytics strategy is easier said than done. We all want to feel like our efforts helped to propel our business forward, but efforts cannot be quantified without understanding performance first. Below is a set of steps that guides the development of a strong digital measurement strategy.

1. Select KPIs that align to your business and digital objectives. 

Before choosing KPIs, ensure that your business and digital objectives are specific and concise. These conditions form the foundation of a strong measurement and analytics strategy. The more vague your objectives are, the more difficult it will be to identify appropriate KPIs and deliver results against them.

Once you have a good handle on your business objectives, choose the metrics that should serve as KPIs for each digital objective. The model below illustrates this concept.

Consider the following basic example. Assume that you work for Company A, which sells widgets online and in stores. Your main business objective is to increase revenue. To do so, you’d like to increase revenue via digital channels. An example KPI could be revenue via digital channels.

2. Define goals for each KPI and report against them. 

KPIs need context; you should know where you’ve been and where you want to go. Benchmarks and goals anchor KPIs.

Let’s return to the Company A example. The revenue goal for this year is $10,000,000. According to Chart 1 we’ve earned $8,000,000 in revenue so far this year. With the assumption that revenue is growing at the same rate each week, we’re tracking ahead of schedule in terms of revenue.

Goals can manifest in a variety of forms. Benchmarks and forecasts can guide them but in the absence of this data, a range of expected outcomes may be used instead. For example, based on variables including (but not limited to) level of competitiveness, monthly search volume, position, and quality score, we can expect a range of paid search click-through rates (CTR). Considering the range of CTRs and landing page elements,  we are able to anticipate a range of conversions via paid search.

Based on the level of sophistication, goals can also be as simple or as complex as needed. In their simplest form, goals could be target volumes (e.g. we want to earn 100 social shares on our widgets site). Or, in a more advanced form, goals could be configured as an index based on a custom algorithm (e.g. we want to earn a widget engagement index of 120 this year).

3. Use analytics to inform digital marketing decisions. 

So your KPIs are aligned to your business objectives and you’ve articulated educated and ambitious goals for these KPIs. Now it’s time to apply your understanding of performance toward making optimizations and decisions.

Generally, basic metrics can assist with making tactical decisions while more sophisticated metrics can inform strategic decisions. For example, email sign-up conversion rate could inform decisions about the layout and calls to action on the page. At a more advanced level, email sign-ups could be assigned a financial value; when this data is compared with the financial data for other conversions (or traffic sources or media channels, etc.), it can guide investment decisions.

To demonstrate, Chart 1 tells us how well we’re tracking against our revenue goal but it doesn’t tell us the full story. On Chart 1, we can see that paid search drives over half of digital revenue but it doesn’t show us the spend efficiency for each driver. Let’s direct our attention to Chart 2 below.

In the past 6 weeks, return on ad spend (ROAS) for paid search and display has been becoming more similar. Ceteris paribus, in the future we may want to test reallocating some funds from paid search to display to understand and optimize the impact on revenue. Or, at minimum we should continue to monitor ROAS by driver to see if the pattern continues and make optimizations when more data is available. To reiterate, this is a very basic example meant for illustration only. It goes without saying that many more factors could impact a decision to test the reallocation of funds.

From a digital cross-channel perspective, the scope of web analytics should cover each area of the marketing funnel: awareness at the top, consideration and engagement in the middle, and purchase intent closer to the end of the path to purchase. This scope could include performance data from the following areas:

• Website behavior
• Organic search
• Paid search
• Social media
• Display media
• Blogger outreach
• Email/CRM

The more you know about performance, the better position you are in to make strong marketing investment decisions. To make your analytics program even more meaningful, aggregate it with data from traditional channels as well.

In conclusion, narrowing your focus to metrics that correspond to business objectives, defining goals for these metrics, and using the results to make marketing decisions can ensure your web analytics strategy is actionable.

Google continued to emphasize high-quality sites and relevant results with updates to the Panda algorithm and page quality. There were additional edits geared toward better detecting natural language, promoting user-friendly results and providing more information on search engine results pages (SERP).

Included below is a graph of the various algorithm updates (categorized by type) from the past 5 months, as well as key highlights from the June / July edits:

Edits to the Panda algorithm

• Description: Google refreshed, updated and added new data to the Panda high-quality sites algorithm. In July, Panda was launched for Japan and Korea.
• Impact: Google continues to emphasize high-quality sites to provide users with the most relevant results.

Multiple updates related to clustering of web results

• Description: Google continues their efforts to make clustering of web results better and more simple.
• Impact: Users are more likely to see multiple results from one domain, potentially benefitting authoritative brands.

Three updates to Image search results

• Description: Updates to signal more relevant image results and filtering of small, unhelpful images to bottom of search results page.
• Impact: Users are more apt to find salient, relevant images. Focusing SEO efforts on all site assets could lead to additional exposure.

Several updates to the display of Sitelinks on the results page

• Description: Google has made changes to make Sitelinks, removing boiler plate text in Sitelink titles, instead emphasizing more useful text.
• Impact: Actively monitoring Sitelinks to demote, in addition to logical site architecture, may result in more Sitelink visibility.

Google added detailed information for Olympic queries

• Description: For queries related to the Olympics, users were provided with detailed information, such as schedules, medal counts, events, and records.
• Impact: Users are able to retrieve Olympic information without having to click on an external site.

Hover feature allows users to quickly compare local places

• Description: When hovering over local places in search results, information appears on the right side of the page.
• Impact: Information provided allows users to compare local places without having to click on an external site. Local SEO may take a more prominent role in personalized results.

Updates focusing on natural language detection for search queries

• Description: Updates were made to better understand natural language for multiple user queries (sunset/sunrise feature, MLB queries, dictionary search feature, etc.).
• Impact: Google continues their commitment to better understand users’ queries and provide more information on the SERP.

First, here are some SEO key differences between Google and Baidu based on their importance:

Source: http://www.codethat.co.uk/blog/search/baidu-seo-guide

As you can see, there are both similarities and differences in the weight each Search Engine gives to the ranking of a Website. One big difference is with the optimization of Meta Keywords. Google does not factor Meta Keywords into its algorithm to ensure keyword stuffing is no longer an influence in page ranking.

Host Servers
Multiple sources have confirmed that it is extremely important to host your site in the area you are targeting. Baidu puts a heavy weight on whether or not the site is hosted locally.Basically, if your goal is to rank highly in Mainland China, then your site should be hosted in Mainland China.

Title and Meta Tags
Title Tags are very important for both Google and Baidu. Page titles need to reflect page content while taking into account the overall number of search queries for the keywords. Meta Description is not important on Google’s overall ranking, however, it is still an essential part in attaining higher click through rates. On Baidu, both Meta Keywords and the Meta Description need to be optimized with relevant, high performing keywords because this is used in their algorithm for higher site ranking.

On-page Content
Like Google, your Website copy should include the relevant keywords you have identified as important to drive traffic. It is also important to pay attention to the right keyword density when choosing keywords. Multiple sources have recommended 6-12% for Baidu optimization.

Linking
A major difference between Baidu and Google is Baidu’s lack of technology to determine link quality. The site gives more weight to the quantity of your links compared to quality. It is still important to incorporate keywords in the internal anchor text.

Content Language
With the many dialects in China, it is important to note which one to choose when building the site. Since Baidu is developed in Mainland China, it is recommended to use simplified Chinese with keywords and on-page content.

Alt Tag
Alt tag with keywords incorporated into Alt text is good for Baidu optimization. However, it is not advised to stuff too many keywords inside.

Baidu is constantly improving their algorithm to compete with the relevancy that Google and other Search Engines provide. It will take some time for their engineers to update the technology that will be able to match the results that Google is able to provide. However, they are making strides and could be there sometime in the future. It would be beneficial to apply the same techniques for Baidu as you would for Google, because the future algorithm update might be coming sooner than you think.

Among May’s announcements were increased presence of Google+ pages on SERPs, dynamic replacement of page titles on SERPs in cases of excessive character counts, and enhancements in Google’s focus on fresh content.

Below is a graph of the different algorithm changes (categorized by type) from the past 3 months, as well as May’s search quality highlights as they pertain to online marketing:

Three updates related to the importance of content freshness

• Description: Continued commitment to Google’s “freshness” update from November [link] – 3 general updates to the algorithm in regard to ranking fresh content
• Impact: Content is gold to Google, and assuming the quality is up to standards, blogging (while implementing proper SEO tactics) can only help

Better recognition of queries intent on discovering fresh content

• Description: Google is getting much better at understanding when searches may be aided by recently published content
• Impact: Similar to above, but this relates more specifically to search terms

Updates to Google+ pages in right-hand panel

• Description: Google+ pages will start becoming increasingly prevalent in the preview panel on the right-hand side of Google SERPs
• Impact: Brands should take note that the increasing prevalence of Google+ in search means increasing relevance

Alternative title may now be generated and appear on Google when meta title is too long or irrelevant

• Description: Page titles that are excessive in character length will revert to a shorter, Google-generated title on SERPs if no custom title tag is provided
• Impact: It is more important now that page titles be shorter than 70 characters, and custom title tags should always be provided and should be succinct

Better detection of “inorganic backlinks” (fabricated spammy links)

• Description: Enhanced detection of spammy link-building techniques designed to try to fool Google’s ranking algorithm
• Impact: Fewer shortcuts for less-established competition who rely on SEO manipulation

Further updates to April’s “Penguin” algorithm

• Description: Heightened emphasis on sites using “white hat” SEO and demotion of those attempting “black hat webspam” loopholes
• Impact: Related to above – active punishment for competitors who rely on SEO manipulation, such as keyword stuffing in anchor text

Didn’t we all agree about 8 years ago that pop-ups were the worst part of the internet? I guess some of us missed that meeting. And somewhere along the way, companies were convinced that they had to use an additional and intrusive analytics platform to gather insights about their users.

This is the first of a four-part series on using web analytics to arrive at the same user insights that pop-up surveys provide. The most basic of these surveys seek to understand how users engage with the site, what they think about layout, how intuitive they found the navigation, did they find what they were looking for, etc. Most, if not all, of these questions relate to user experience. And most can be answered by looking at existing data.

First, to make sure we are on the same page, let’s take a look at some of the surveys. I don’t mean to pick on Forsee (there are plenty of companies offering this type of measurement), but this was the first one I came across – on the Forsee site.

I don’t want to get into the usability of the survey itself, but should’t they at least tell me how many questions it is or how long it will take? Also, I think the argument could be made that there is way too much text there for me to make an informed split-second decision about what to do. In any event, clicking ‘Yes’ does something even worse than a pop-up…it triggers a pop-under.

Once I leave the site, the pop-under turns into the survey. A 22 question, 2,400+ pixel tall survey.

We all know how the surveys work. Let’s look at the survey question by question and find alternative less-intrusive ways of gathering the same data.

Question 4: Please rate the visual appeal of this site
Question 5: Please rate the balance of graphics and text on this site
Question 6: Please rate the readability of the pages on this site

This is where technographics come into play. By analyzing user behavior segmented by browser, operating system, screen resolution, mobile vs. desktop, etc., trends may begin to appear that indicate poor aesthetics, layout, readability, etc. For example, we will often create several custom segments that only display data for users on various versions of IE. In doing so, we are able to isolate content that may not appear correctly due to IE 6-8 bugs. We will also look at page load time across these technographic details to ensure that content is loading relatively quickly for all platforms, browsers, etc.

Certain behaviors on a site can also be an indicator that the content is difficult to read. If a site is using a button to allow users to enlarge text or increase contrast, those actions should be tracked and attributed back to technographic data. If the data suggests that certain browsers are enlarging text, for example, the styling of the content can be adjusted for that specific browser to be more readable.

Question 7: Please rate how well the site is organized
Question 8: Please rate the options available for navigating this site
Question 9: Please rate how well the site layout helps you find what you are looking for
Question 10: Please rate the number of clicks to get where you want on this site

There is a great feature within Google Analytics (and other analytics platforms) that allows you to see the path that users take through your site. You can segment by geography, browser/OS, traffic source, users who convert, etc. and view the number/type of pages they viewed during their visit. I won’t go into too much detail related to the specific tool, suffice it to say that this report can help to answer the questions above. For example, how many times do users have to click before they perform a specific action? Do users seem disoriented (e.g., do they keep going back to the home page or repeat the same navigation loops)? Are users who convert cycling through all of the main navigation before they find what they are looking for, or is it a direct path?

Question 11: What is your overall satisfaction with the ForeSee website
Question 12: How well does the ForeSee website meet your expectations

Metrics such as bounce rate and conversion (segmented by traffic source and user type) would do a better job of answering these questions. This brings up a larger issue that I have with these surveys, which is that I think the results are skewed, and only coming from those on either end of the ‘user experience’ spectrum. Only those with truly horrible or fantastic experiences will likely take the time to complete a 22 question survey. I think the actual user behavior metrics gathered for every user provides a more reliable and representative data set.

In addition, overall satisfaction can also be measured using social sharing metrics, on-page engagement (e.g., commenting, printing, etc.) pages per visit, visit frequency/recency and all of the other metrics we associate with a good user experience.

Question 13: How well does foresee.com compare to your idea of an ideal website

I’m not quite sure of the purpose of this question. Or how you act on the answers. Regardless of the score, I would find myself asking, ‘So what? How do I improve or maintain these scores?’ The other questions seem to do a fair job of answering this question in a more specific and actionable way.

Question 14: How likely are you to return to the ForeSee website

We don’t need users to tell us whether they will return or not. We can look at our site analytics and see whether they actually do. And similar to some of the other questions, the ability to segment new vs. returning visitors by traffic source and user type makes this data point much more useful than a user’s feelings right after leaving the site.

Question 15: How likely are you to recommend foresee.com to someone else

Do you have social sharing buttons on every page of your site? Are you measuring that activity in your analytics platform? If so, you know which traffic sources, content types, user types, etc. are more likely to share or recommend your product.

Question 16: How likely are you to request additional information about ForeSee products or services in the next 30 days

Cookie users who visit your site but do not convert. Set the cookie to expire after 30 days and track whether those users come back and convert (e.g., fill out a form, request more info, download a free trial, etc.). By analyzing these conversion rates, the time lag between the various visits and actual conversion as well as segmenting by traffic source and content viewed, you can begin to understand the consumer journey for each user type. Once you know average conversion rates, you know how likely various user types are to convert within 30 days – and you can begin optimizing that journey.

Question 17: How frequently do you visit foresee.com

Every decent analytics platform allows you to see how frequently users visit your site. Even better is the ability to segment this data by traffic source and content viewed – to see which marketing efforts, content and user types seem to be repeat visitors.

Question 18: What is your role in visiting this site today

In this question, they are hoping to identify whether the user is a current client, employee, student, etc. This is easily discovered by analyzing navigation paths and content viewed. If someone repeatedly returns to the site via a direct visit, logs in, signs up for training or visits the site from an email sent to current customers, they are probably current customers. If a user visits twice per month to read the latest news releases and then bounces, they are probably a news/media organization. If they are watching demos after arriving from a non-branded search, they are probably prospective customers. Setting up advanced segments and custom reports to measure the differences between user types will tell you more than just who the users are, but what they are looking for, how they found you, etc.

Question 19: What were you looking for on the ForeSee website today

Search query data, whether from organic or paid search, can yield insights into what users are looking for. Measuring site search queries (i.e., users utilizing the search function on your site) can also provide this data. Interestingly, it can also be a good indicator of content that the user found difficult to find – especially if it is used in combination with navigation paths. For example, if someone arrives at your site, visits several pages, goes back to the home page, starts over, repeats this behavior, and ultimately uses the site search, their query answered this question for you.

Exit rate isn’t always a bad thing. Outside of search data, exit rates can also be an indication that a user found what they were looking for, particularly if it lines up nicely with their traffic source, search query, navigation path, etc.

Question 20: Please tell us the industry in which you work, if applicable

I’ll give them this one. I’m not sure how to replicate this data in a web analytics specific platform unless you have a self-identifying navigation (i.e., ‘I am a doctor’ vs ‘I am a patient’) or content geared towards specifically industries. On the Forsee site, for example, they do have content and white papers that are industry-specific. Setting up an advanced segment for each industry that identifies users that seek out and share that content could provide insights into user behavior, traffic source, etc. for users within that industry. By doing this, you can not only identify which industries you appear to be reaching (i.e., number of pageviews by industry segment), but also the efficacy of that content (i.e., conversions after having viewed that content).

Question 21: Please share with us your suggestions to improve the ForeSee website

Isn’t this function replaced by a contact us form or side-mounted feedback button (similar to those provided by Get Satisfaction and UserVoice)?

Question 22: Would you like a ForeSee representative to contact you

Again, solved with a simple contact form.

In summary, most of the questions above are aimed at gathering feedback related to user experience (ironically, the vehicle itself is a bad user experience). And most of the answers can be found by looking at the data currently being gathered by a decent analytics platform. I wonder if users who answer question #10 include the click to make the popup go away in their total.

You probably noticed that I left out questions #1-#3.

Question 1: Please rate the accuracy of information on this site
Question 2: Please rate the quality of information on this site
Question 3: Please rate the freshness of content on this site

In fairness, I’m not sure how to replicate these questions using web analytics. Calculating the accuracy and quality of content could be measured relative to other user behavior such as social sharing, conversion rates, return visitation, etc. However, that correlation seems a bit questionable.

I welcome arguments from both sides. Do you think the value of direct user feedback outweighs the detriment to user experience? Do you think that users who complete these surveys are representative of average users?

In parts 2, 3 and 4 of this series I’ll discuss other types of surveys designed to gather data related to brand awareness, brand affinity and purchase intent – and how we can also measure these KPIs without direct surveys.

Consolidation of paginated content

• Description: Search results will tend to consolidate multiple pages of an article into a single listing.
• Impact: Using rel=”next” and rel=”prev” tags can help index paginated content and consolidate authority around a topic.

More domain diversity in search results

• Description: Rather than many results from the same domain, SERP’s will feature more diverse domains.
• Impact: Ranking for non-branded queries may be easier with more diverse domain results.

Search snippets show more text from the beginning of pages

• Description: Search results are likely to show text from the beginning of a page, if relevant.
• Impact: Feature relevant content with priority keywords near the beginning of pages.

No ‘freshness boost’ for low-quality content

• Description: The promotion of ‘fresh’ web material now excludes content identified as low quality.
• Impact: Even if fresh & new, low quality content will not rank.

More concise & informative titles

• Description: When reviewing factors related to titles, Google may adjust this element to be more informative or concise.
• Impact: Create relevant and concise page titles so Google shows your preferred version, rather than generating its own title.

More spelling corrections globally & for longer queries

• Description: Longer queries are now included for detailed spelling corrections; global impact.
• Impact: Optimizing for mis-spellings is a strategy that will not benefit for auto-corrected queries.

Expanded sitelinks in search results

• Description: Expanded sitelinks now scrutinized by individual page ranking & refreshed often.
• Impact: Maintain high quality category pages & update xml site maps dynamically to reflect content.

Keyword stuffing detector improvements

• Description: Improved detection of pages that are using keyword stuffing.
• Impact: Keyword usage should be natural and relevant.

So What is the Penguin Update?

This Google update was unique in its lack of details. Perhaps the SEO community has been spoiled lately with detailed algorithm factors (at a high level) given that years past have been much more black box. The impact to site ranking is just starting to become apparent and shared information among webmasters has revealed a few negative factors Penguin targeted:

• Demotion of low-quality exact-match domains (e.g. dna-paternity-test.net, paternitytest4u.com, etc.)
• Aggressive exact-match anchor text (found in footer links and other internal links)
• Links from low-quality sites (indicating an effort to manipulate search results)
• Sites that “spam” Google with keyword stuffing and cloaking (serving up different content to Googlebot than a visitor receives)

Let the White Hats Rejoice!

The fun part about Google cleaning house is that when they demote the web derelicts, SERP’s are able to pull in more quality content and authoritative sites that have been doing “white hat SEO” for years. These are the kind of tactics that Barefoot Proximity promotes; transparent site activity, natural link building and minimal attempts to manipulate content to suit Googlebot. Sadly, the white hats have been forced to watch as low-quality spam sites have outrank them for years, benefiting from many of the tactics that Penguin has explicitly targeted in this update.

Reviewing organic search traffic (and Webmaster Tools data when available) for a handful of large-scale properties reveals a small bump in organic search traffic around April 24th and 25th. These examples represent different industries (CPG, insurance, advice content) and, because of this, are all the more indicative of a Penguin impact. We did not see this uptick universally across all our sites and time will tell if the increase is temporary as the SERP’s normalize or representative of a permanent shift.

Over-Optimization and the Future of SEO

While the tone here reflects the optimism of a new day, let’s temper our excitement by remembering what prompted Penguin in the first place. SEO’s always try to optimize their sites to the 10th degree (we are a perfectionist bunch) and in doing so possibly fall into the trap of what Matt Cutts recently referred to as “over optimization”. While that phrase might give search marketers the jitters, a big caveat here is that a moderate level of typical SEO behavior is still appropriate and beneficial. Matt references those “black hat” SEO’s that take site efforts so far as to buy inbound links, tweak anchor text to the point of detracting from user experience and generally try to manipulate Google in a behind-the-scenes manner. Barefoot Proximity has never believed that such tactics are worthwhile for a long-term search strategy.

Between the Panda and Penguin updates, Google has trained their crosshairs on low-quality spam sites that bring little joy to the searchers of the world, helping promote white hat efforts while discouraging the alternative. The ominous warning about “over optimization” is something to take note, though perhaps the semantics are implying more than Matt intended. This much is still true about SEO:

• Sites should continue promoting microdata markup
• Clean semantic code helps Google crawl and index sites faster
• Fast site speed is becoming a new mandatory
• Author credibility will increase as a ranking signal
• Other web best practices, adoption of which Barefoot Proximity is committed to promoting for their clients

As demonstrated in the screenshots above, updates like Penguin designed to clean up the web allow quality and authoritative sites to take their rightful place in the SERP’s.

The below updates represent only a small percentage of algorithm changes made, with Google only publicly acknowledging 35 changes though many more are likely left unmentioned. The items below are worth noting either in their urgency to take action or reinforcement of current SEO best practices.

Sitelink Data Refresh

• Description: Sitelinks are generated automatically from site structure & internal links. This process will now happen more frequently.
• Impact: Ensure site maps clearly indicate site architecture. Internal links should include relevant and descriptive anchor text.

Improvements to image search relevant

• Description: Google signals better promoted “reasonably sized” images on high-quality landing pages.
• Impact: Compressing high file size images will better fit Google’s definition of “reasonably sized”.

Better app ranking and UI on mobile devices

• Description: When searching for apps on a mobile device, rich results with icons, ratings and a download button will be formatted for mobile displays.
• Impact: Adoption of microdata, both on applications described here and page elements in general, continues to gain prominence in the search algorithm.

Improvements to freshness in video results

• Description: Google can better detect ‘stale videos’ and instead serve up fresh video content.
• Impact: Video xml sitemaps & microdata markup provide publishing dates and other relevant meta data.

Better handling of queries with navigational & local intent

• Description: Queries with both local & navigational intent (seeking a specific website) receive more relevant results.
• Impact: Creating localized landing pages specific to your domain/brand could receive additional listings in personalized results.

Improvements to name detection

• Description: Google has updated its system to recognize personal names, including celebrities.
• Impact: If your brand has the endorsement or partnership of a celebrity, his/her name should be utilized across digital assets.

More precise detection of old pages

• Description: New signals identify “stale” content, serving those pages less in search results.
• Impact: Publishing frequent content is always ideal, though updating old content in meaningful ways can avoid a “stale” designation.

+1 button available in more countries and domains

• Description: The +1 button gains a more international profile.
• Impact: Including +1 functionality can improve global search traffic.

ayout of a page in an attempt to gauge user experience.

Above the Fold

When we think about user experience we aim to make it as painless as possible for users to find, and engage with, the content they are looking for. By placing huge expanding ads at the top of a page or filling the header and sidebars with unnecessary information, CTAs, ads, etc. we are hurting that experience. And now Google is confirming that those types of pages will have a harder time appearing in organic search results. Great news for user experience fanatics like us. Bad news for sites trying to milk every pageview for as much ad revenue as possible and sites that push real content lower on the page than is necessary.

Curious what sections of your site’s pages appear above the fold? Google has a browser size tool that overlays the percentage of users that will see the various sections of your page.

SEO & Design

This block level analysis emphasizes the need for an SEO strategy to sync up with and, in some ways, inform the wire framing and design of a website. It is also another indication that Google is getting better at calculating user experience and beginning to reward sites with good layouts, fast loading pages, easy/intuitive navigation, accessible/responsive design, etc.

At the end of the day, Google wants the best possible experience for their users – and so should we.

Rand Fishkin recently discussed the accuracy of visitor data for SEOMoz site traffic and the sometimes egregiously incorrect data provided by these tools. He rallied a call-to-arms for Webmasters to share the level of deviation on their own sites and form an industry consensus. We took Rand’s spirited dissertation to heart and analyzed a large (anonymous) site, comparing actual data with the public estimates.

Alexa.com – Lots of Data (Most of It Wrong)

Alexa gets credit for offering more data than other free public services, including search keyword traffic, clickstream data and audience demographics. The only issue being the extremely inaccurate nature of this information.

The top queries from search traffic were not indicative of actual organic search activity, with a few queries not appearing at all in the real data. Perhaps most inexcusable was the “% of traffic from US”, comparing Alexa’s 74% with the actual 96%. This is an important metric to advertisers and the discrepancy here could very much impact our business opportunities.

The demographic of site visitors has been shown to be Males, 35-44, (confirmed through a number of methods) but Alexa interpreted our visitors instead as Females, 45-54, which is the exact opposite from reality.

Compete.com – Under-reporting Visitors & Proud of It

Compete reported Unique Visitors at 57.92% lower than actual data, though some penance is given with the option to export CSV data. They might be forgiven a margin of error for sudden swing months, though this large discrepancy is averaged over 7 months time.

Actual Data: 4,596,476 unique visitors

Compete Data: 1,934,332 unique visitors

DoubleClick – Google Gets It Wrong

You would expect the master of data to have accurate numbers, especially considering that the crux of their revenue stream depends on adoption of AdWords and AdSense. The margin of error is not much different than Compete and Alexa at 63% under-reported Unique Visitors and 24% under-reported Pageviews.

Unique Visitors:

• DoubleClick: 68,000
• Actual: 185,733
• Difference: 63.39%

Pageviews:

• DoubleClick: 830,000
• Actual: 1,086,206
• Difference: 23.59%

Quantcast – Great if Quantified (Otherwise Not)

Electing to place a Quantcast tag on your site (hereafter known as being Quantified!) allows you to publically share real data. The tag does not estimate, nor does it guess, but instead provides accurate and reliable data. However, not everyone is comfortable with this level of transparency and instead relies on Quantcast to estimate their traffic. A quick review of 3 large sites shows unique estimates under-reported by 85% – 90%. Not even close to reality. Similar to Alexa’s missteps on audience demographic, Quantcast viewed our Males, 35-44, site instead as Females, 50+.

ComScore – You Pay for the Best of the Worst

The only paid service mentioned here is noticeably more accurate than the alternatives, showing only a 12% pageview under-reporting discrepancy compared with real data. Being the service most widely respected, it was refreshing to see ComScore actually coming close to providing true analytics.

Actual Data: 10,161,527 pageviews

ComScore Data: 13,519,743 pageviews

The goal of these traffic tools is to offer a public face for your site without giving away all that valuable data. So in that context estimations are perfectly useful, though a margin of error between 10%-15% is not an unreasonable request for general accuracy. All tools error on the side of conservative under-reporting, which is an understandable tendency, though the degree of discrepancy should still fall within moderate boundaries. It is interesting, however, that many tools show a similar magnitude of error around 60% under-reported, perhaps indicate that the same (inaccurate) data source is referenced by multiple tools.

A recent research study by Ogilvy Action quantifies both the increase in mobile prominence and also its effect on digital shopping habits. Noting that “mobile search traffic doubled in 2010 and has gone up five-fold in the past two years”, the study targets Singapore as the model of early adoption for mobile technology and a bell weather for global trends to come. The chart below demonstrates the great potential for global growth if the current pace of adoption continues or accelerates.

The role of mobile search was particularly prominent during the recent holiday shopping season. In a Nov 2011 Google study regarding toy shopping, it is reported that “nearly 1 in 5 toy shoppers use mobile to shop and search is their #1 activity”, followed by visiting mass merchant and retailer sites. Paid mobile search ads are perfectly targeted to capture the intention of online toy shoppers while they are actively searching. The role of mobile devices is becoming more of a “shopping assistant” during a consumer’s purchase process. The chart below from another Google/IPSOS study reveals the extent to which shopping decisions are influenced by mobile devices.

Retailers with physical stores are noticing the trend and adopting tactics that compliment mobile habits. This past holiday season some Lowes stores provided employees with iPhones to better interact with shoppers who were scanning product bar codes or searching for online reviews.

Mobile phones are becoming ingratiated within all aspects of our lives. Brands that recognize this trend and position themselves within the mobile space – through a mobile optimized website, custom phone application or appearance within search – will reap the benefits of increasingly savvy shoppers and their mobile phone “shopping assistants”.