http://metia.com/seattle/olga-spaic Wed, 24 Apr 2013 23:48:00 +0100 Wed, 24 Apr 2013 23:48:00 +0100 Olga Spaic /seattle/olga-spaic/2013/04/impact-of-new-ftc-digital-advertising-guidelines/#number-of-comments The Federal Trade Commission (FTC) recently updated its guidelines for disclosures in digital advertising.  Most of these updates will have the biggest impact on social media advertising where space for disclosures is often limited.  Advertisers were always required to disclose any caveats around their marketing offers, now they just have to incorporate the new consumer reality – customers on-the-go, scanning mobile devices and not having the time to read small print, multiple disclosure pages, or research the validity of ads or claims by a celebrity endorser. Bottom line:  The rules have not changed, the technology has.  Read more in my article in Seattle 24x7.  /seattle/olga-spaic/2013/04/impact-of-new-ftc-digital-advertising-guidelines/ /seattle/olga-spaic/2013/04/impact-of-new-ftc-digital-advertising-guidelines/ Wed, 24 Apr 2013 23:48:00 +0100 Olga Spaic /seattle/olga-spaic/2013/04/authenticity-is-key-to-social-media-success/#number-of-comments Customer engagement in an authentic voice is the key to success in social media. Whether you have an entire team supporting your social presence or a lone part-time marketer, that authentic voice is what convinces customers that they should care about your brand (and that your brand cares about them). So when resources are tight and shortcuts via automation seem like a more efficient solution, think about the risks of this approach and its impact on your customer relationships. Will automated tweets alienate your customer base? Will inappropriately scheduled Facebook posts suddenly land you in hot water with your fan base? Will you be perceived as insensitive by continuing with business-as-usual on social channels in light of a national tragedy? Gartner’s 2013 US Digital Marketing Spending Survey provides an interesting perspective on authenticity in social media: One of the biggest – and most underappreciated – challenges of social marketing is achieving consistent, authentic brand communication at scale. 47% of our survey respondents see content creation and curation as the most important role of their social marketing teams (more than any other category), yet they struggle to come up with a consistent stream of relevant and compelling “on-brand” messages. So they look to agencies for help. Content and social engagement tops the list of ways that marketing service providers are helping clients with social marketing. Let’s take a look at a few examples where engaging with a digital marketing agency like Metia might help a company mitigate risks in the social media realm and drive “on-brand” authentic conversations with their customers. Consistent on-brand messaging requires forethought and planning.  Many companies are eager to join the social media melee but don’t put forth the necessary resources and time to develop a strategy and plan for their engagement.  This plan needs to take into account other marketing efforts the company is undertaking.  A content editorial calendar can help coordinate cross-channel efforts.  One of our clients, the AT&T Developer Program, relies on an all-up editorial calendar to drive communications via email, blogs, Twitter, Facebook, and in-person events with the developer community.  If there was ever an audience that demanded authenticity in social media, it’s the developer community. The teams at AT&T and Metia have worked together for several years developing relevant messaging to the developer community, shifting away from overt marketing messages to more technical and relevant content this audience could be excited about.  These efforts led to a 113% increase in Facebook fans, a 136% increase in Twitter followers, and a 16 point increase in their Klout score during 2012. You can read more about the AT&T Developer Program coordinated social media efforts in this case study on MarketingSherpa. Automation and prescheduling posts can create snafus for companies. Take a look at Fox TV Network, which has an extensive social media presence for all of their network shows. One of their shows, Terra Nova, was pre-empted by the World Series Game 5 in 2011. This meant that scheduled tweets about that night’s show were no longer accurate, but they still went out because they were prescheduled. This created confusion among viewers as they tried to figure out if there was still a broadcast – and, meanwhile, there was no response from the Fox network.   Breaking news and tragedies often create confusion for corporate brands. Is it better to turn off all communications or to show support for those impacted by the tragedy in your messages? Take a look at the tweets from Epicurious earlier this week after the Boston Marathon tragedy:  Needless to say, these were not received well by the Twitterverse as they came off as completely insensitive and self-serving.   We offer some lessons for marketers looking to create that authentic voice: (1)    Always keep an eye on your social media posts and scheduled emails. Is there breaking news that might make your messaging seem insensitive, offensive, or perhaps just outdated? a.       If yes, pause all planned messages until you understand the situation better. Only resume after a respectable time period (typically the next day) and proceed with caution.  b.      Be careful not to post self-serving messages in the wake of a tragedy.  (2)    Avoid auto-responses. It’s better not to acknowledge a new follower than to send an automated tweet with a generic link. Social media is about personal connections, and some customers might be dissatisfied with auto-replies. Unfortunately, dissatisfied customers are more likely to share their unhappiness broadly than those who are happy. (3)    Like your other marketing channels, social media content needs to be engaging. Ask questions, solicit feedback, respond to queries from your followers – these are all ways to build that personal connection and engage with your customers. These tips should be a part of your broader social media policies and employee and agency guidelines. These guidelines should also speak to your corporate voice and values, as well as outline any specific standards those engaging on your company’s behalf in the social media realm should abide by. /seattle/olga-spaic/2013/04/authenticity-is-key-to-social-media-success/ /seattle/olga-spaic/2013/04/authenticity-is-key-to-social-media-success/ Fri, 19 Apr 2013 18:11:00 +0100 Olga Spaic /seattle/olga-spaic/2013/03/industry-benchmarks/#number-of-comments One of the things that we are often asked when talking to clients is how they are performing vis-à-vis industry benchmarks.  Is their website on par with competitors?  Are people as likely to click through their emails as those of other programs?  I often have to caveat any response by saying that “it depends” – each website and email is unique in its purpose and how it engages users, so industry averages or benchmarks are very rough comparison points.  Research companies (comScore, Forrester, Nielsen, and others) will often publish reports which include this type of data, but most of these services require the user to actually purchase the report or subscribe to their services.  Many marketers have limited budgets and cannot afford these reports, but here are a few “free” resources that you can take a look at in case you are searching for benchmarks. The Fireclick index Fireclick’s KPIs are based on client data (all proprietary) and tend to be primarily for companies with an e-commerce focus.  The index has existed since 2003, and users can look at data for the past 7 days, 30 days, 6 months or 1 year. MarketingSherpa Another great resource is Marketing Sherpa.  Some of the more recent gems they posted include a chart with average website conversion rates by industry:   … and a mobile email click-through and open rate chart:   MarketingSherpa publishes a number of market research studies throughout the year – those are not free, but they will often publish excerpts from the reports on their website.  Subscribe to MarketingSherpa’s email newsletter to get updates on those reports and other articles of interest.   In the web arena, one of the most sought after metrics is the Bounce Rate.  This one seems hard to come by, but Anil Batra, an analytics blogger, did some of his own research in 2008 to come up with bounce rate benchmarks.     Keep in mind that anything in the digital marketing space changes very quickly.  Thus benchmark numbers that are more than a couple of years old should be used with caution – you cannot assume that your CTR will be as good as those seen 5 or 10 years ago when users were not quite as bombarded with content. Depending on how much data you can access, you can also create your own benchmarks – just focus on comparable email campaigns or websites, and see how your KPIs compare. If all else fails, keep tabs on trends in your data over time to see if your digital marketing initiatives are making an impact on your scorecard metrics. /seattle/olga-spaic/2013/03/industry-benchmarks/ /seattle/olga-spaic/2013/03/industry-benchmarks/ Fri, 08 Mar 2013 20:33:00 +0100 Olga Spaic /seattle/olga-spaic/2013/02/big-data-and-scientific-method/#number-of-comments Rarely has a business term caused as much excitement and confusion as the buzz-worthy big data.  Big data promises to deliver great insights about your customers - nearly every technology and consulting company is talking about it as the holy grail for marketing (oops, another buzz word!).  But what do you as a marketer really need to do to make the most of big data? Read my article on tackling big data with the scientific method on DestinationCRM.com.     And let me know what you think! /seattle/olga-spaic/2013/02/big-data-and-scientific-method/ /seattle/olga-spaic/2013/02/big-data-and-scientific-method/ Thu, 21 Feb 2013 22:26:00 +0100 Olga Spaic /seattle/olga-spaic/2012/07/even-at-the-olympics,-data-is-tops/#number-of-comments All eyes have been focused on the Olympics this past week.  And one of the most commons stats seen across media outlets is the latest medal count – which country will ultimately win the most gold medals?   Behind the scenes though, it’s all about the user experience and understanding how spectators both at the events and on media channels, are responding to the Games.  This involves analyzing an astounding amount of data… an analyst’s nirvana!  The goal, of course, is to make the Olympic Games the very best event for all of those customers. One of the largest worldwide research organizations, Nielsen, has been running research studies for the London Organizing Committee of the Olympic Games for the past three years.  Their research has influenced the choice of mascot, the ticket pricing mechanism, the organization of the volunteer corps at the Olympic Park, and the execution of the closing ceremony.  During the games, Nielsen deployed the latest technology and research techniques, including mobile and app-based surveys, to measure spectator sentiment.  Nielsen’s research efforts have been so successful and chockfull of insights that organizers of future Olympics have already reached out to Nielsen for assistance.  You can read more about some of these research initiatives here. An even more fascinating look at the data is NBC’s so-called Billion Dollar Olympics Lab.  Ok, maybe the research lab didn’t cost a billion dollars (though NBC’s broadcasting rights did), but the reams of information being gathered by the media provider is astounding.  Just think about this:  the network is broadcasting 835 hours of programming on 6 different channels, over 3,000 hours of live-streaming on their Olympics website and over 3,200 hours via mobile and tablet channels.  Every single user visit, click and comment is being tracked and analyzed on NBC’s websites as well as via their partners comScore and Google who are capturing data from user panels. One of the more interesting aspects of this research is a look at the so-called “simultaneous media user” - as mobile and tablet technologies have developed over the past few Olympic Games, the NBC research team has seen a significant impact of these users on overall viewing behavior and engagement.  This is especially important as the digital media channels are often preferred by the younger generations, who were previously thought of as uninterested in the Olympics. Of course, much of NBC’s research is also driven by a desire to demonstrate ROI to its advertisers.  One of the examples noted from the Vancouver Games was a spike in Google search traffic immediately after a British Columbia tourism ad – a pretty obvious correlation of the impact of advertising on consumer behavior!  More details on the Billion Dollar Olympics Lab can be found here. So even though NBC has received their share of criticism over its “delayed” broadcasts of events over the past few days, the hope is that all of this great data and research will inform future decisions and result in a win-win-win – for NBC, the spectators and the Olympians. /seattle/olga-spaic/2012/07/even-at-the-olympics,-data-is-tops/ /seattle/olga-spaic/2012/07/even-at-the-olympics,-data-is-tops/ Tue, 31 Jul 2012 01:20:00 +0100 Olga Spaic /seattle/olga-spaic/2012/06/marketing-and-big-data-–-still-about-the-basics/#number-of-comments How can big data help companies become more customer-centric?  At the Wharton Leadership Forum yesterday in Seattle, Wharton School Professor Peter Fader and Jarvis Bowers from Microsoft’s IEB unit discussed the various ways companies are using data to segment their customers and provide unique, value-added experiences.  Though many companies collect data on customers, very few know what to do with it – and when they think they do, they tend to oversimplify it by slotting customers into generic segments.  Does it matter that a 20 year old and an 80 year old belong to different demographic groups if their observed behavior is the same?  This is just one of the questions brought up during the discussion.  The conversation was hopping and covered a wide range of topics, but I jotted down a few takeaways.  No statistics background required – this is marketing common-sense. Customer-centricity is not about doing everything for every customer.  It’s about knowing which customer is more valuable and perhaps treating those customers just slightly better than the rest.  Professor Fader pointed out that Starbucks and Nordstrom’s, known for their great customer service, roll out a “generic red carpet” and make everyone feel good for stopping into their locales. But he says, being “nice” is not a competitive advantage – it can be easily copied.  Instead, Fader suggests trying to attain a clear understanding of the customer and their value to the company as a way to build its competitive advantage.  In that vein, he mentioned a new Neiman Marcus app that alerts store staff when customers come into the store – presumably, high-value customers would sound the alarm quickly and get immediate attention from staff.  So much for anonymous browsing!  On the other hand, Fader thought that Amazon.com does a great job of maximizing revenue based on customer behavior and is much more selective with its customer perks.  Bowers, however, countered by saying that Amazon doesn’t always get a 1-to-1 relationship – in the instance of their household for example, all four family members use the same Amazon Prime membership.  The parents and kids have individual preferences that are not necessarily reflected in this household-to-Amazon relationship.  I would contend that Amazon can still parse that data to determine the specific preference of the household members – a children’s toy or book is obviously something that is a preference of a child in the household, just like a business book is likely to be a preferred item of Bowers himself. If you ask customers for data, do something with it.  Bowers noted that many companies start loyalty programs but don’t actually provide any valuable information or perks to those “loyal” customers.  Customers expect a company to engage with them after they have provided an email address.  One of the more interesting examples of a clear value proposition early in the Internet era was ESPN.com – avid sports fans had to create online profiles to participate in the site’s fantasy sports leagues.  ESPN.com provided a forum for engagement, all the while collecting reams of data on their customers – both the information they volunteered in their profiles as well as clickstream data on their behavior.  This was a win-win for both the users of the site and ESPN.  Fader pointed out that loyalty programs can’t be “if you build it, they will come” – they really need to expand upon your company’s value proposition and look for ways to engage customers so they can be repeat customers, influencers of others in their purchases, and so on.  In today’s Internet era, loyalty programs may not have as many rules and a very rigid structure but they still need to follow this concept of “engagement” – if you put up a Facebook page, don’t just focus on getting Likes, but also spend time on engaging with those fans. You don’t need all imaginable data, just the data you will use to make smarter business decisions.  Big data  presents a unique problem for many companies – instead of thinking strategically about the data they need, many companies try to collect as much data as they can and then try to figure out what to do with it.  Instead, both Fader and Bowers recommend thinking about data more strategically – what do you really need to do know to make smarter decisions? This is especially important in the context of privacy and personally identifiable information (PII) – customers expect something in return in exchange for giving their PII, and they expect you to use their data judiciously.  So the less you ask for, the easier it is to be a good privacy steward too! Figuring out customer lifetime value (CLV) is worth the effort.  Knowing the CLV and the aggregate customer equity (sum of all of the CLVs of your various customers) can help you prioritize your company’s investments over the long term.  Many companies tend to talk a lot about CRM, 1-to-1 marketing, and using data to determine customer lifetime value, but very few actually take on the hard work of creating the right models to truly measure CLV.  Bowers said that this is paramount for companies in ultra-competitive spaces as they need to extract value from every customer interaction – it is not a “nice to have” for them. There was no shortage of examples from the real-world business community – the  discussion touched on a number of other different brands including American Express, Apple, Orbitz.com, Tesco, Paris Casinos, and Bonobos.  Both the academic Peter Fader and the practitioner Jarvis Bowers had a similar parting thought – focus on blockbuster customers not blockbuster products.  Use the data you have on customer behaviors to figure out how to both maximize and enhance the relationship with your customers and extract more value from those relationships. /seattle/olga-spaic/2012/06/marketing-and-big-data-–-still-about-the-basics/ /seattle/olga-spaic/2012/06/marketing-and-big-data-–-still-about-the-basics/ Fri, 29 Jun 2012 22:25:00 +0100 Olga Spaic /seattle/olga-spaic/2012/05/maximize-roi-with-testing/#number-of-comments In a blog post a few weeks ago, Amber Whiteman wrote about Testing for email communications.  Testing is actually important for all digital communications, not just email.  You can test anything –background colors of pages, images, call-to-action links or buttons, price points of products you might be selling, and so on.  Anything you are not sure about, you can test.  Whatever your goals might be for your digital marketing efforts, you should test to determine if your current site has the engagement and response rates you expect, and if not, then use the test results to optimize content and improve that performance.  It’s less costly and less risky to test ahead of a big campaign than to skip an A/B test and go down a possibly unsuccessful path.   As an example, let’s say you are testing price sensitivity of your customers for a specific product. You might show 50% of the site visitors a price point of $10, and the other 50% a price of $10.99 for product A.  Your initial hypothesis might be that the visitors who see the $10 price point will be more likely to purchase, but the test might also reveal something unexpected.   Perhaps referral data (search engines, other sites, etc.) indicates that certain keywords or websites referring visitors drove sales of the product at the higher vs. lower price. Or the less price sensitive customers (those who made the higher price purchases) were more likely to buy additional items along with product A.  Or maybe those preferring the lower price were quicker to change their minds and abandon their shopping carts?    Though your initial hypothesis might be that the lower price might directly correlate to a higher engagement or click-through rate, the higher price might be resulting in a broader array of other purchases which could ultimately be resulting in higher overall sales on your site.  Depending on the results of the test, you would adjust your marketing efforts accordingly and drive greater overall profitability for your business.   As a practical matter, a simple A/B test is relatively straightforward – as Amber explained in her post on email testing, you would send a portion of your site traffic to one web page/site treatment and the remainder to the other treatment.  You can run this test at the same time and portion off traffic, or run each scenario sequentially (first option during one week, second option during the subsequent week, etc.)    However, to save time and money, and to give you the opportunity to test lots of different options including perhaps some off-the-wall ones, you might want to undertake multivariate testing where you test multiple treatments at the same time.  What if you were to test 2 background colors, with 4 different price points, 2 purchase buttons and 4 product images?  There are actually 64 possible combinations to test.  A sampling of what the combinations and results might look like is in the table below. If you were to test each of these combinations separately, it might take several weeks to do so and yet you have a campaign you want to launch as soon as possible. A multivariate test, for all of these different scenarios would allow you to quickly determine which combination is the “winner” and optimize your website accordingly to take advantage of say a new influx of traffic to your site from a recent advertising campaign.  In the sample above, the “optimal” combination based solely on conversation rate (note, we only have 8 of the 64 listed) looks to be number 8.   There are many tools for conducting multivariate tests (Google’s Website Optimizer, Webtrends Optimize, SiteSpect, LiveBall, Visual Website Optimizer are just a few) – these testing platforms automatically split off your website traffic to run through the various scenarios.  Many of these platforms allow you to run tests without tinkering around with your website code – these make it much easier for marketers to run tests without getting support from your engineering team.  Once the test is complete, the tools often integrate directly with your web analytics package to provide you with easy to decipher results on which options perform the best.  No need to be a statistics buff to minimize risk and optimize your pages!   There are many tools for conducting multivariate tests (Google’s Website Optimizer, Webtrends Optimize, SiteSpect, LiveBall, Visual Website Optimizer are just a few) – these testing platforms automatically split off your website traffic to run through the various scenarios.  Many of these platforms allow you to run tests without tinkering around with your website code – these make it much easier for marketers to run tests without getting support from your engineering team.  Once the test is complete, the tools often integrate directly with your web analytics package to provide you with easy to decipher results on which options perform the best.  No need to be a statistics buff to minimize risk and optimize your pages! /seattle/olga-spaic/2012/05/maximize-roi-with-testing/ /seattle/olga-spaic/2012/05/maximize-roi-with-testing/ Thu, 31 May 2012 14:32:00 +0100 Olga Spaic /seattle/olga-spaic/2012/02/mind-your-data-google-updates-its-privacy-policy/#number-of-comments It seems that controversy over online privacy is unlikely to subside any time in the near future.  Last week, Google announced that it would be simplifying its privacy policies across their entire service offering starting March 1, 2012.  All Google service users received alerts and emails urging them to read up on the changes as they would be consolidating 60 different policies into one shorter and easier-to-read policy.  Transparency is a key tenet of a good privacy policy as I wrote a few months ago in Target Marketing, so I was excited to hear about this change. But being transparent has landed Google amidst controversy yet again, as several privacy-focused Congressmen requested additional information in a formal letter to Google’s CEO.  The battleground intensified this week as Microsoft launched print ads bashing Google’s use of customers’ data purely for their own benefit.  Here’s a snippet from the ad (click on the image for more): At the bottom of the ad, Microsoft encouraged users to tap into its own service offering – Hotmail, Bing, Office 365 and Internet Explorer which are purportedly different from Google in how they collect and use your data.  Google’s public policy department responded with their own blog post refuting these claims. The problem with all of this brouhaha is not Google’s privacy policy itself – the new privacy policy clearly states that Google wants to unify the customer experience and unify your data, which it may and will use to further improve that experience by providing you, the end-user, with more relevant content.  Some of that more relevant content comes in the form of advertising, which is after all, how Google can offer a free service and still make money.  If you are interested in what type of demographic data Google has inferred about you, take a look here: https://www.google.com/settings/ads/onweb/.  In my case, since I tend to block tracking in my IE browser, Google does not have a very detailed view into my online persona. What’s changing with the new privacy policy is that this type of profile will be developed by streamlining and connecting data across many Google services, which will likely mean a more detailed picture of each end-user. This is what Google called out as the biggest change in its policy: The main change is for users with Google Accounts. Our new Privacy Policy makes clear that, if you’re signed in, we may combine information you've provided from one service with information from other services. In short, we’ll treat you as a single user across all our products, which will mean a simpler, more intuitive Google experience. Google goes on to explain that you can opt out of targeted advertising and customize your privacy settings in the Google dashboard (though not all of it without completely foregoing some major Google services).  The FAQs also provide relatively simple definitions of common terms related to online tracking and profiling, though I think the jargon is still too techy for less savvy Internet users. The lesson here is that all Internet users should be aware that data is collected about them by Google, Facebook, Microsoft, and many other online service providers, and that they use this data to serve up more targeted content.  Simpler, more transparent policies allow you to more easily understand that what you do on the Internet can be tracked, and what you can do to protect your data and your online persona.    Regardless of the website you are visiting, you can also adjust your browser settings to limit tracking.  Most importantly, you can impose a self-filter on your own online behavior and avoid posting sensitive information online without first understanding how it can be used by the websites you are visiting and their third-party advertising partners.  In this world of instant social media sharing, this may be a hard thing to do, but it’s really the only way to maintain your privacy. /seattle/olga-spaic/2012/02/mind-your-data-google-updates-its-privacy-policy/ /seattle/olga-spaic/2012/02/mind-your-data-google-updates-its-privacy-policy/ Thu, 02 Feb 2012 20:03:00 +0100 Olga Spaic /seattle/olga-spaic/2011/09/social-media-metrics-roi-or-engagement/#number-of-comments I came across an interesting article today from eMarketer entitled: “As Social Spending Rises, Which Metrics Are CMOs Focusing On?“  Naturally, my ears immediately perked up as this is part of an ongoing conversation among analysts in the digital marketing community.  For several years now, we have been hearing the question – what’s the ROI from social media? The eMarketer article, which summarizes a CMO survey spearheaded by Duke University’s Fuqua School of Business and the American Marketing Association points to different priorities and not ROI – namely customer relationship and engagement metrics.  What happened to ROI?  One would think in a slow-to-grow economy that ROI would still be top dog as marketers look to prioritize investments.   But instead, simpler metrics like hits, visits, page view, repeat visits and number of followers or friends are the ones at the top of the list. In one sense, I agree with eMarketer’s conclusion that marketers are focused on building customer relationships which will over time prove their value as more loyal customers drive greater demand and revenue.  However, I also think there is another reason – ROI-based metrics are hard to get!  Many companies deal with multiple analytics tools and many sales tracking databases.  The customer is likely exposed to marketing and sales activities from multiple channels – traditional media, online ads (whether they be search, banner, email, social media), retail displays, newsletters, social media feeds, calls from sales people, and so on.  On the surface, online channels seem like they would be better at tracking individual users and how they are responding to a specific product offer – but a simple choice by an online user to go into a retail store to buy the product instead of clicking through to an online purchase can impact even those relatively clear-cut online success metrics.  Imagine then, what happens when you look at the cost of marketing across all channels on a per customer basis (if it can even be quantified) and the impact of those marketing dollars on ROI.  Until there is a simpler solution and a way to connect all of those dots, marketers in a sense don’t have a choice – they have to focus on the simpler metrics because that’s all that they can get.   /seattle/olga-spaic/2011/09/social-media-metrics-roi-or-engagement/ /seattle/olga-spaic/2011/09/social-media-metrics-roi-or-engagement/ Fri, 16 Sep 2011 00:23:00 +0100 Olga Spaic /seattle/olga-spaic/2011/05/metrics-for-marketers-conversion-rate/#number-of-comments I find myself often answering questions with other questions.  For example, say a client asks: - What’s our conversion rate? - Where? - On our website? - Which part of the website? - The whole thing. - Hmmm, what are the goals of your site?  Have we set up tracking for your specific calls-to-action? Conversion rates are one of those key performance indicators (KPIs) that everyone working in digital marketing has heard of – and a good conversion rate is the ultimate goal.  But let’s step back and think about what a conversion rate actually means.   Of course, being the prolific Internet user that I am, my immediate go-to source is Wikipedia:     - Great.  So what’s our conversion rate? - the client reiterates. Let’s look at a simple example of my blog page.     What would I like my readers to do? A - Click on an article title B - Contact me via email C - Follow me on Twitter D - Share the article page E - Subscribe to my blog RSS The above would be my list of key calls-to-action (CTAs), or in Wikipedia speak, my goal achievements.  In most web analytics packages, I can track the clicks on these links and calculate my conversion rate from the blog home page.     Assuming other Metia bloggers have similar goals, we can calculate an overall Metia blog conversion rate.  What about the rest of the site?  First, I would caution against coming up with conversion rates for each page on a site – some pages are not visited frequently and are probably not pages that you necessarily need to drive users toward.  In the case of our site, most of the site consists of blog posts, so focusing on blogs is just one approach.  But our HR manager, for example, is probably most interested in how many job applicants her blog posts (job listings) are yielding – so in her case, she would probably most care about clicks on the email link.  What’s nice about tracking conversion rates for her section is that she can actually compare how the responses to job listings on our site compare to those on other sites and determine where conversion rates are highest.  Now if only the sites could help her determine the quality of the job applicants… but analytics can only go so far.  If you work in a larger company that has a broad reaching site, there will likely be many sections of the site that you are probably not going to care about as a marketer.  For example, a section on technical support – you certainly want your customers to find that information if they need it, but it’s probably not the content you would be driving them to if you are trying to sell them products or services, right?  Your focus is most likely a campaign landing page and the campaign CTAs – a buy now button, a link to register for an event, etc. There is, of course, debate on how useful it is to calculate a site or page conversion rate.  If your site is not selling actual products, can you measure the impact of improvements in your conversion rate easily?  And if you invest in improving that conversion rate, will the investment pay off?  For an e-commerce site, for example, you will likely want to be more focused on revenue-specific KPIs.  For a support site, you will likely want to focus on metrics that help determine the effectiveness of your online support content (e.g. customer satisfaction or content ratings) and how that content is helping to diminish actual calls to your call centers.  Certainly, all of these types of activities can be used to calculate a conversion rate, but just be mindful of the fact that the conversion rate on the support site should probably not be compared to a conversion rate on a marketing site. Bottom line: If you choose to measure conversion rate, think about your site goals and what you are trying to achieve.  And then once you identify those goals, determine what metrics are the ones you can impact – if you can impact conversion rate and it helps you improve the business, then by all means, include it in your web analysis efforts.   /seattle/olga-spaic/2011/05/metrics-for-marketers-conversion-rate/ /seattle/olga-spaic/2011/05/metrics-for-marketers-conversion-rate/ Wed, 25 May 2011 23:57:00 +0100 Olga Spaic /seattle/olga-spaic/2011/04/metrics-for-marketers-bounce-rate/#number-of-comments In my day-to-day conversations with marketers, I’m often asked about the perfect metric or KPI for measuring marketing effectiveness.  There is no simple answer, however, as a given metric is not one-size-fits-all.  Every marketer has slightly different marketing objectives and thus needs to keep track of a slightly different set of metrics. In this weekly series, I’ll highlight some of the more popular and useful metrics that marketers should keep an eye on.  I’ll focus on metrics that are not your typical measures of traffic to a site – traffic measures are great indicators of overall awareness and reach, but are often not the ones where an analyst can find nuggets of insight.  If you happen to have questions on specific metrics, please post them in your comments and I’ll address them in future blog posts. This week we begin with Bounce Rate.  According to the definition on Wikipedia, a “bounce occurs when a web site visitor only views a single page on a website, that is, the visitor leaves a site without visiting any other pages before a specified session-timeout occurs.”   A visitor can “bounce” from the page for a number of reasons – clicking on a link that takes them to a different website, closing the window or tab on which they were viewing your site, typing in a new URL, going back within their browsing session, or because their visit simply timed out (typically 30 minutes within the web analytics industry).  As a marketer, why should you care about bounce rates?  A bounce rate is an easy way to determine if your site (or pages on your site) is engaging visitors.  If your site is a one-page landing page or a portal that drives your users to other sites, then a high bounce rate is probably a great result.  However, if you are trying to get users to interact with the site, learn about your products and services, test out demos, view videos, Tweet about cool articles, and so on, then a high bounce rate means that your users are not finding your site or page compelling enough to even warrant a click.  Yikes!  At a high level, you should at least track your site’s overall bounce rate, but it’s even more insightful to get into the nitty-gritty – bounce rates for specific high-traffic pages (Top 10 pages), top entry or referring pages, specific campaigns including SEM, and so on.  Looking at bounce rates at this more granular level can help you identify potential problem areas on your site that are often easy fixes. As an example, this is what a bounce rate report looks like in WebTrends, a popular web analytics tool. Is 71.9% a high bounce rate?  According to Avinash Kaishuk, an analytics guru from Google, “it is hard to get a bounce rate under 20%. Anything over 35% is a cause for concern and anything above 50% is worrying.”  We won’t name the site here, but in this instance, the site does in fact act as a portal – many users land on the page from other industry-leading sites or search engines, read the relevant article and then leave or look for more details on the product/service they just read about on another part of the domain (which is not tracked as part of this specific WebTrends profile).  Thus, bounce rate over 50% is not a major concern. However, on a different site where the overall bounce rate is lower than the previous example, we should probably examine the details.  Engagement with key calls-to-action is one of the main objectives of this site and a bounce rate of over 60% indicates that visitors are either landing on the site unintentionally or not finding the pages they land on compelling enough to click further. Unfortunately, going into the details of the analysis of this site would reveal too much proprietary information, but rest assured, the site owners pay attention to the bounce rate and several other engagement metrics in their efforts to improve the user experience and overall satisfaction with the site. I’ll focus on a different metric next week - remember to comment if you have one you are particularly curious about. /seattle/olga-spaic/2011/04/metrics-for-marketers-bounce-rate/ /seattle/olga-spaic/2011/04/metrics-for-marketers-bounce-rate/ Tue, 26 Apr 2011 05:43:00 +0100 Olga Spaic /seattle/olga-spaic/2011/02/not-all-data-aggregation-is-bad-consumers-benefit-from-real-estate-market-analyses/#number-of-comments Being a bit geeky about data, I was intrigued by the headline a few weeks on Mashable.com: Who Owns Your Data?  The conclusion of the article is that the actual ownership of data is irrelevant but it’s what you do with the data – that is, how you analyze it to draw out insights and make it more valuable.  This is something I think about every day in the business context, but lately, I have actually been thinking about data ownership, aggregation and analysis from a more personal perspective – the point of view of a homebuyer as my family embarks on a search for a new house. Up until a few years ago, accessing real estate information in the U.S. was a very cumbersome process – the so-called Multiple Listing Service (MLS) was only available to realtors and thus, even for a simple I’m just curious what’s on the market query, you either had to engage a realtor or drive around a neighborhood hoping to spot a “for sale” or “open house” sign.  Fast-forward to today.  Real estate data is virtually everywhere and it’s nicely packaged on a number of different websites.  These websites provide aggregated information from both the public and private domains - tax assessment records, transfer of deeds, recent comparable sales, existing MLS listings, community ratings of schools, and various other disparate sources to make house-hunting a lot easier.  Two of these sites, which I happen to be using for our house-hunt are Zillow.com and Redfin.com, and luckily since they are headquartered right here in Seattle, their data on this market are particularly robust. Zillow’s core differentiator is its so-called Zestimate® – it provides a buyer or seller an estimated market price of a property given all of the data inputs and current market conditions.  The site has interactive charts with pricing trends, maps that clearly list estimated values of other neighboring houses, links to nearby comparable sales, and of course a whole series of sponsored links from realtors, lenders and a few Google ads, for good measure.  If you own a property, you can subscribe to alerts about your property’s estimated value, indicate a “make me move” price, and update the information on Zillow with details about remodels or upgrades you have made, which will hopefully result in a higher Zestimate® for your property.   For a homebuyer, all of the Zillow data and analysis is an absolute dream – if you find a house you like, you are no longer making an offer blindly, but can make an informed decision and hopefully save a significant amount of money!   Redfin.com operates a little differently from Zillow.com in that it primarily showcases houses actually for sale.  In addition to typical listing information, a buyer can get details about how long the home has been on the market and changes in list price over time (invaluable in getting a sense of how flexible the owner is on the list price), ratings of nearby schools, statistics about recent comparable sales, and links to those oh-so-useful Zestimates®.  What I most like about Redfin.com is its iPhone/iPad app – the interface is simple and it’s meant for the house hunter on the go.  It immediately identifies the nearest home for sale, allows you to search for homes by neighborhood or by zooming in on a map, and when you find the place of interest, you can view online photos and all of the details found on the main website, as well as get Google Maps directions from your current location to the house.  If you snap a photo or jot down a few notes, the info synchs up with what you might have saved in your online search results so you can access everything when you get home and wrap up your house-hunting research.  You can also set up alerts/searches to ping you whenever a new listing comes up in your target area – unlike other sites, you are not required to rely on postal codes or town names but on the actual mapped area you find of most interest.   I absolutely love having this type of information at my fingertips – and I love the fact that someone else did the legwork and aggregated and analyzed everything for me.  It’s certainly not perfect, but it sure beats going into the housing market completely blind and being completely reliant on a single real estate agent for all of your information.  Armed with the data and analysis provided by these two sites, I’m sure to score a deal on our future house. /seattle/olga-spaic/2011/02/not-all-data-aggregation-is-bad-consumers-benefit-from-real-estate-market-analyses/ /seattle/olga-spaic/2011/02/not-all-data-aggregation-is-bad-consumers-benefit-from-real-estate-market-analyses/ Fri, 25 Feb 2011 00:40:00 +0100 Olga Spaic /seattle/olga-spaic/2011/02/rooting-for-watson/#number-of-comments Last week, fans of the game show Jeopardy were treated to a three-day competition between IBM’s latest supercomputer – Watson – and two of the show’s best known champions, Ken Jennings and Brad Rutter.  I found myself conflicted as I watched the show.  I started rooting for the super computer as it’s truly a marvel of human engineering and shows remarkable progress in the field of artificial intelligence, but then I was disappointed when Watson would take over and not let the super-smart humans even chime in with a response… After all, if the computer wins, does it mean that computers are now smarter than humans? As it turns out, even though Watson won the show (and hopefully that wasn’t a spoiler), it still isn’t smarter than some of the smartest humans.  Watson was designed to perform well within the particular scenario of a Jeopardy game where, within 3 seconds, it could process its database of various topics to come to a reasonable response (actually, several responses ranked by statistical likelihood of being correct) and click on the response buzzer.  The fast processing also allowed Watson to determine “contextual” relevance of the responses and, in some sense, to understand the nuances of natural human language.  But for many questions, the speed to click on the buzzer was really all that Watson had as an advantage as the human competitors also knew the answers.  In several instances, Watson was stumped and, in fact, at the end of the first game, it totally botched up the Final Jeopardy response by guessing a Canadian city for the category U.S. cities. Listening to a post-game analysis of the show on NPR’s “To the Point”, I learned about something called a Turing test.  This is a test of a machine’s ability to demonstrate human intelligence.  It was conceived by the British mathematician Alan Turing in 1950 (yes, more than half a century ago).  During this five minute test, a human judge is tasked with posing questions to two participants (one is a human, the other a computer) and needs to determine by the end of the test which of the two respondents is the computer.  Over the years, this test has been run countless times and though computers are getting smarter, they are still not able to imitate human conversation to such a level of sophistication to fool the human judge.  The IBM scientist who worked on Watson conceded that Watson would not be able to pass the Turing test.  So is Watson only good for game show playing?  Shortly after the Jeopardy publicity stunt, IBM announced some of its first agreements to deploy Watson in the real-world – one of them is the medical sphere where a Watson-like computer can act as a physician’s assistant.   For example, a patient with a specific medical issue comes in to see the doctor.  While in the examining room, the doctor can pose questions to the patient to determine a possible diagnosis.  Meanwhile, Watson would be listening in, and providing additional inputs for the doctor.  Watson can easily pull up and process the patient’s medical records, as well as the latest medical and pharmaceutical research that the doctor might not have had time to read, and present those insights to the doctor for additional consideration as he forms his final diagnosis.  Sounds like a very useful application of supercomputing power! Of course, I then began wondering if the world of marketing analytics would one day benefit from computers like Watson.  Certainly, much of our industry already relies on mass computing power to capture clickstream data, process it, and render it in pretty charts and graphs.   What, of course, is missing and where the human element comes in is to determine what the data means and if a change to a specific banner ad, an update in SEO keywords, or a link on Twitter caused an improvement in traffic and conversion rates.   We spend a lot of time testing these changes and inputs, but sometimes we don’t have a complete view into all of the inputs… so we re-test hypotheses, tweak websites, and see what the data tells us after all those tweaks.  But wouldn’t it be really cool to use faster processing power of a supercomputer to look at multiple data streams quickly and help us reach insights sooner to dazzle our clients?  Perhaps, Watson and super smart computers aren’t all that bad. /seattle/olga-spaic/2011/02/rooting-for-watson/ /seattle/olga-spaic/2011/02/rooting-for-watson/ Mon, 21 Feb 2011 02:30:00 +0100 Olga Spaic /seattle/olga-spaic/2011/02/targeted-ads-annoyed-web-users/#number-of-comments Over the past several months, there has been a “raging” debate in the U.S. over targeted ads – some of this stems from lack of knowledge among users, some of it stems from the continued discussion among government agencies on privacy policies and how to regulate use of data, and some of it is just a bit of journalistic sensationalism as stories about young users of Facebook being bullied (or FourSquare users being stalked) abound. I thought an article from eMarketer today on the issue of targeting was particularly insightful. As the dust begins to settle on the privacy debate, it seems that users are eager to adopt some sort of “do not contact” solution much like the “do not call registry” which was instituted in the U.S. in 2004 to remove consumers from telemarketers’ phone lists. Already some web browsers are offering users the ability to click a “do not track” button, but the question remains if the industry will adopt and honor these preferences? Or will the government have to get involved to manage that adoption? And how does this play out in the international arena?  Plenty of questions to ponder. /seattle/olga-spaic/2011/02/targeted-ads-annoyed-web-users/ /seattle/olga-spaic/2011/02/targeted-ads-annoyed-web-users/ Tue, 08 Feb 2011 05:56:00 +0100 Olga Spaic /seattle/olga-spaic/2011/01/why-i-love-apis-and-a-cool-way-to-play-with-all-that-data/#number-of-comments As a web analyst, one of the challenges of my job is dealing with disparate data sources – very frequently, instead of having access to one report that I can slice and dice any which way in a web analytics reporting solution, I have to dig through countless disconnected reports and figure out a way to connect them all so I can paint the story of what web users might be doing in response to a particular campaign or a change in the navigational structure of a website.  Sometimes this can be fun if you put on your “detective” hat, but at other times, there is just too much data and not enough time in the week to clean it all up and organize it into consumable bundles.  Besides, most clients are much more appreciative when you spend your time on data analysis to draw insights and recommendations – even the smallest nugget of insight (which might seem insignificant to you as an analyst) can make them smile. Thus, one of my favorite features built into most popular web analytics tracking software is the application programming interface, or lovingly referred to, as the API.  With help from our developers, we can set up a whole slew of reports that pull in data from the analytics API as well as other data sources, providing a holistic view into how the business is performing both on and off the web. There will be plenty of opportunity to talk about how to get at insights from data in future blog posts, but I just want to put in a plug here for one of the coolest reporting tools now available as an Excel add-on - Power Pivot.  What’s nice about Power Pivot is that it basically allows a non-SQL expert to manipulate large arrays of data with ease within the much more familiar Excel interface.  I’ve been playing around with it recently and there are some awesome features, one of which just happens to be leveraging the new Windows Azure Data Market which is loaded with tons of free data sets with demographic data, census data, tax data, and so on. There are data sets that cost money too, but just think of the wealth of information available in the public domain that is now at your fingertips and easily accessible using PowerPivot…  that is what I call analyst nirvana!        /seattle/olga-spaic/2011/01/why-i-love-apis-and-a-cool-way-to-play-with-all-that-data/ /seattle/olga-spaic/2011/01/why-i-love-apis-and-a-cool-way-to-play-with-all-that-data/ Mon, 31 Jan 2011 05:34:00 +0100