jump to navigation

See you in LA next week? April 1, 2008

Posted by Elana Anderson in Customer Analytics, Customer Experience, Database Marketing, Integrated Marketing, Marketing, Marketing Measurement, Marketing Strategy, Marketing Technology, Online Marketing.
Tags: , , ,
add a comment

Busy, busy, busy! That’s the month of April for me. Sorry I haven’t been posting the last couple of weeks, but I have a ton of balls in the air at the moment. One of the major things I’ve been working on is a minor treatise aimed to help marketers systematically improve the relevance of their customer communications. My work on this whitepaper, “Marketing Beyond The Status Quo,” is sponsored by the good people at Responsys and we’re planning to unveil it together next week at Forrester’s Marketing Forum in Los Angeles.  Our session is Tuesday, April 8th at 2:25pm. I’ll be presenting along with Scott Olrich (Responsys’ CMO) and Randy Cuff (Director of CRM Development at Intrawest). Here’s the abstract for the session:

Few interactive marketers are ecstatic about their ability to deliver meaningful and timely marketing messages. In fact, most agree that more relevant and timely marketing communications will be better received by customers and increase response rates. However, when time is scarce, budgets are tight, and single channel campaign management solutions are already integrated and delivering ROI, marketers are hard pressed to change the status quo. In this thought-provoking session, Elana Anderson, former Vice President and Research Director leading Forrester’s marketing practice, and Scott Olrich, CMO of Responsys will unveil the findings from a first-ever study identifying the drivers of the “status quo” paradigm, and reveal the strategies and marketing technologies smart marketers are using to deliver superior marketing performance and ROI across channels.  

If you are attending the forum, please stop by our session. I’ll be there for the full two days, so drop me an email if you want to connect at some point during the event. Hope to see you there!

Advertisement

Recalibrating The Meaning of “Relevant” March 11, 2008

Posted by Elana Anderson in Customer Experience, Database Marketing, Integrated Marketing, Marketing, Marketing Measurement, Online Marketing.
Tags: , , , , , , ,
3 comments

Have you ever refinanced your mortgage only to be bombarded by offers to lower your mortgage rate for six weeks after you closed on your new loan? This happens because financial services companies are purchasing “triggers” from credit bureaus that indicate you have had a recent loan approval. The problem with these triggers is that they are not timely. By the time the marketing communication gets to the customer, it’s too late.

Relevance = right message + right time + right place

Perhaps it is a cliché, but it’s a good one. Too many marketers focus entirely on the message component of relevance. For these marketers, “place” is typically an outbound channel and “time” is based on the internal campaign calendar – not the customer’s needs. To be relevant, marketers must step outside the confines of the functional silo that they are responsible for and think outside in – from the perspective of the customer. In addition to targeting the message itself based upon a customer’s stated or implied needs, relevance requires:  

  • Timely reaction or response to customer actions. Although some marketers are experimenting with trigger-based communications and on-site customized messaging, the prominence of these tactics pales in comparison to the weekly or semi-weekly campaign pushes. The beauty of these tactics, however, is that they can be automated.
  • Cross-channel integration. Your customers don’t care that you are only responsible for email and not the website or direct mail and not the call center. When a potential customer clicks on a search result or an online ad and lands on your generic home page or receives an offer in the mail and calls customer service to inquire further, he expects a seamless handoff. Yet, creative elements often dominate conversations about integrated marketing rather than a focus on what the customer is trying to achieve as he traverses the channels. The result of this oversight? For the customer, it often means dead ends and unnecessary frustration. For the company, it means lost opportunities and, possibly, damage to the brand.
  • A programmatic approach. Did you know that maximizing individual campaign response might be to the detriment of overall program ROI? That’s right. Sending more messages may generate a higher response, but how many others are simply tuning out? The current industry standard in the retail sector is 1-2 email messages a week. Amazon differentiates itself in the inbox by not always being there. An email offer from Amazon might be, “A brand new Leonard Cohen CD is available. Since you have enjoyed Leonard Cohen in the past, we thought you might want to know…” This programmatic approach requires different metrics than the campaign-centric approach – for example, program engagement over time or revenue per customer (not campaign).

Demonstrating The Brand Value Of Email January 29, 2008

Posted by Elana Anderson in Customer Analytics, Database Marketing, Marketing Measurement, Marketing Strategy, Online Marketing.
Tags: , , , , ,
1 comment so far

I had the opportunity to catch up with Robert Raines, VP of Product Management & Creative Services at E-LOAN last week. I first met Robert several years ago when he was getting the company’s email programs off the ground. Robert shared with me some of the things that he’s accomplished since we last spoke and, as always, he had some great insights.

E-LOAN wanted to evaluate the impact of its email program 

Robert is a firm believer in testing and, more specifically, using control groups to measure the impact of email marketing activities. To determine the long-term benefit of its email program, the E-LOAN team created a randomly selected universal control group. While the control group still received transactional email communications (e.g., “We have received your application”), it received no email marketing treatment whatsoever for a period of 18 months.

To ensure that individuals selected into the control group were excluded from all email marketing efforts, the team created an exclusion table in its marketing database and automatically excluded the control group from any database extracts that were sent to its email service provider (Responsys).

What E-LOAN learned

What was E-LOAN looking for? The company wanted to evaluate the application rate of the mailed population vs. that of the unmailed population (the control group) over an 18-month period. As you might expect, at the beginning of the test there was very little difference in the application rate of the two groups. But, over time, the emailed population had a significantly higher application rate. So much higher that, according to Robert, the difference alone is enough to justify the total annual cost of the company’s email program.

Robert is also quick to point out that it’s not just about being in the inbox that matters. The E-LOAN team works hard to ensure that its email program is relevant and it uses a mixed strategy that includes broadcast messages (e.g., “The Fed has lowered interest rates”) and highly targeted, event-triggered communications.

Recommendations

Email marketers often complain that they don’t have enough staff and struggle to manage what’s already on their plate given the staff that they have. I believe that this complaint becomes a self-fulfilling prophecy… Email is cheap, just blast it out, and we’ll keep the bare bones operation going… To break through this cycle, email marketing managers should:

  • Devote 6-8 hours of the team’s time to implementing a test and measurement strategy. As a manager myself, I know that it’s possible to squeeze some amount of extra time out of the week. If you are really committed to making email marketing more strategic in your company, find 15-20% of someone’s time and focus it on test and measurement as Robert has done at E-LOAN. Sure, the E-LOAN test that I’ve shared here took 18 months, so start with something smaller. For instance, test the difference between a broadcast newsletter and a newsletter with dynamically targeted content.
  • Trend results over time. A quick test to show that targeted content generates higher open and click rates is indeed interesting. But, it’s more interesting to trend this information over time to evaluate the sustained value of a targeted vs. untargeted program.
  • Document a business case. If your goal is to improve the internal stature of your email efforts, get more budget, and grow your team, then it’s imperative to document your case. Avoid doing this at an individual campaign level and comparing metrics — like opens and clicks — against industry averages. Focus on the bigger picture and build a case that exposes the real business value of your efforts. What is the ROI of targeted vs. broadcast communications? Or, as in the E-LOAN example, do your customers buy more if they receive email marketing communications from you? This is the kind of case that your bosses need to free up more resources.

What To Expect From Your Web Analytics Tool (Web Analytics Series, Part 2) January 15, 2008

Posted by Elana Anderson in Customer Analytics, Database Marketing, Integrated Marketing, Marketing, Marketing Measurement, Marketing Technology, Online Marketing, Web Analytics.
Tags: , , , , , , , , ,
4 comments

In my post just before the holidays, I shared a framework to help marketers think about how web analytics contributes to data-driven marketing effectiveness over time. Marketers agree that they want to deliver more relevant and timely communications, establish a two-way dialog, and generally be more customer-focused and integrated, but many are struggling to make positive progress. Use the framework to understand understand the maturity of your data-driven marketing practices. Then define your objectives and timeframe for making incremental progress. In this post, I want to discuss how marketers can use their placement on the framework to define the key requirements for their web analytics tools. I work with an awful lot of firms out there that are not getting the benefits that they should be out of the technologies that they purchase. Why? Well, I think one key reason is that they are overly aggressive in their expectations of what they will achieve and over what time. The result is a lot of wasted technology — and wasted time.

Web Analytics Maturity Framework

If we were more realistic about what we are trying to achieve (i.e., the business outcome) with our web analytics tools then I believe our learning curve around how to effectively understand and leverage the data would actually accelerate. Why? Because we wouldn’t be constantly struggling with the technology. The free tools on the market are getting better and, while they are not sufficient for marketers that are beyond Stage 1 maturity, they will meet the needs of many. Here is a starting point to help you get beyond vendor eye candy and align functional requirements with business objectives:

Stage 1: Site analysis

Key questions you need to address:

  • How many visitors are coming to my site?
  • How are visitors using my site?
  • How are visitors finding my site?

Core functional requirements:

  • Visitor analysis
  • Referrer analysis (pages and keywords)
  • Strong library of parameterized “out of the box” reports

Comments on the market: These capabilities are table stakes to enter the web analytics market and most of the solutions out there do a reasonably good job here. Expect a more limited library of reports and more limited customization features from the free tools.

Stage 2: Site optimization

 Key questions you need to address:

  • How can I increase site visibility?
  • How do content and taxonomy influence desired action?
  • What would improve site navigation?

Core functional requirements:

  • Path analysis
  • Page and scenario drilldown analysis
  • Drop-off analysis
  • A/B and multivariate testing

Comments on the market: This is currently an area of focus for leading vendors in the market. The once-standalone optimization players – like Offermatica (acquired by Omniture), Kefta (acquired by Acxiom), and Optimost (acquired by Interwoven) – have been acquired and other vendors are looking to add these capabilities through acquisition or organic development. Given current client emphasis on customer experience management, expect this to continue to be an area of hot competition in the near future.

Stage 3: Segment targeting

 Key questions you need to address:

  • How can I logically group site visitors?
  • How can I target visitor content by segment?
  • How can I leverage site learning in other communication?

Core functional requirements:

  • Segmentation model templates
  • Ability to persist segments
  • Ability to create dynamic segments and apply them historically

Comments on the market: Leading web analytics vendors like Coremetrics, Omniture, Unica, and Visual Sciences (acquired by Omniture) offer segmentation capabilities but this is an area where vendors differentiate.

Stage 4: Individual customization

 Key questions you need to address:

  • What is the best content for an individual based on prior site interaction?
  • Should I reach out to an individual customer NOW?

Core functional requirements:

  • Individual visitor profiles retained over time
  • Ability to match profile to current visitor context – in real-time
  • Ability to unify profiles when visitor identifies

Comments on the market: A few leading vendors are really just starting to focus here. Omniture’s recent acquisition of TouchClarity is a good example. Unica is also working on integrating it’s web analytics and campaign management modules in a meaningful way. But, overall, the vendors are just getting started at figuring this out.

Stage 4: Integrated marketing

 Key questions you need to address:

  • How are customers using online and offline channels in the buying process?
  • How can I optimize online and offline interactions?

Core functional requirements:

  • Calculate and retain key profile metrics
  • Track metrics longitudinally
  • Open data model and facilitation of extracts to other systems

Comments on the market: Today, you’re mostly at the mercy of your internal IT shop when it comes to the level of data integration sophistication required to help marketers in large companies integrate their activities across channels. Some firms call upon their interactive agency, systems integrator, or database marketing service provider to help. But, each approach has its challenges. It may make you feel better to know that no one has nailed this one and that gives us all something to aspire towards. At the end of the day, I believe that if the web analytics vendors want to be part of the solution then they need to hire (or partner) with database marketing gurus in order to make real progress.

What Does [Should] A CMO Do? December 18, 2007

Posted by Elana Anderson in Integrated Marketing, Marketing, Marketing Measurement, Marketing Strategy, Marketing Technology.
Tags: , , , ,
1 comment so far

In its quarterly SEC filing, Orbitz announced that it has (also see related AdAge article):

…decided to eliminate the global Chief Marketing Officer position and continue managing the Company’s marketing efforts on a regional basis. In connection with that decision, Randy Wagner, Chief Marketing Officer of the Company, is expected to leave the Company in mid-February 2008.

That’s unfortunate. I’ve met Randy — she was a keynote speaker at the Forrester Marketing Forum last year which I hosted. She’s a bright, strong leader and I’m sorry to see her go. But, I’m more concerned about Orbitz and all of the other companies out there that are cutting CMO positions and/or clearly struggling to define the role.

I think Orbitz is missing the point. I don’t disagree at all with the idea that defining and managing campaigns on a regional level can be a more effective way to drive to growth goals. I just don’t think that the job of the CMO is to define and manage campaigns.

So, what SHOULD a CMO do?

Well, we’ve been talking for years about integrated marketing, customer centricity, customer relationship management, customer experience management, 1:1 marketing, etc… I’ve been intimately focused in this arena myself for a over a dozen years and I feel like we (the broad and royal “we”) have made little progress towards these goals. To be sure, a lot has also changed in the last 12 years but if we EVER want to get there (or even close) then we need a strong leader. And, from my perspective, that leader is the CMO. The role of the CMO should be to:

define and lead a customer-focused marketing strategy that crosses product, channel, geographic, and even functional boundaries.

I realize that this is much easier said than done. It starts with a CEO who believes in the business benefit of being customer-focused and a CMO with the vision, leadership capabilities, and charter to make it happen. It will also require:

  • A complete overhaul of the marketing organization. I’m talking structure and reporting hierarchies, metrics, culture, and process — all of it. What’s the right answer here? Well, as all good consultants say, “it depends”;-) I haven’t found a perfect organization yet. The key is understanding where the organizational weaknesses are and then putting tools or processes in place to help bridge the gaps. First and foremost, however, I believe it starts with the metrics. The CEO, CMO, and CFO need to sit down and figure out how to measure marketing impact in ways that don’t result in marketing teams competing with one another for customer mindshare or quibbling over which team gets credit for customer conversion.
  • Acquiring and nurturing new skills. What skills am I referring to? Left-brained skills: business acumen, process orientation, quantitative analysis, and technical knowhow. We’ve been talking about this one for a while and it is slowly happening. But, marketing leaders often complain that it’s hard to find these skills along with a love of the customer and a passion for marketing all in the same body. I suggest looking for consultants (Accenture, Bain, etc.) who want to get off the road, pillaging your internal IT organization for the systems analysts or project managers that always ask the business questions, or plucking young marketing analysts for whom there is no technical barrier and putting them all in an aggressive mentorship and cross-team training program. 
  • Significant investment in technology and infrastructure. How boring is this one? My POV on this is that rather than going goo-goo gaga over the next trend and treating it as a antidote to all of marketing’s woes, it’s high time for marketing organizations to recognize that technology — and integrated technology at that — is a crucial enabler. To achieve our goals of customer-centric and integrated marketing, we need to manage the marketing process on top of a framework that is, itself, integrated. Again, no easy answers here — there’s no vendor or application out there that will take care of this for you. And, I’m not saying that the CMO has to be a techie. But, a good leader recognizes his/her strengths and weaknesses and surrounds him/herself with a team that can fill the gaps. The bottom line here is that marketing organizations need to have a technology strategy. Those that don’t will NEVER achieve the customer-centric vision or be able to effectively integrate their activities.   

Who is responsible for driving this agenda — on a global level? The CMO.

Now, I also want to be clear here that I was not implying above that the CMO doesn’t have responsibility for the global brand(s)…  She does! Today’s consumers are really good at sniffing out and publicizing inconsistencies between what corporations and their brands say and how they act (think the recent Unilever Dove/Axe controversy). So, today’s CMOs must own aligning every brand under the corporate umbrella with the core values of the corporate entity and reconciling the brands with one another. Companies that fail to do this are at the mercy of the consumer.

Web Analytics Needs To Grow Up (Web analytics series, Part 1) December 11, 2007

Posted by Elana Anderson in Customer Analytics, Customer Experience, Database Marketing, Integrated Marketing, Marketing, Marketing Measurement, Marketing Technology, Online Marketing, Web Analytics.
Tags: , , ,
3 comments

Last week I was privileged to spend the day in the company of several thought leaders in the field of Web Analytics. Dr. Alan Hall, Avinash Kaushik, Judah Phillips, and I participated in several taped panel sessions that focused on how to best leverage web data to improve marketing effectiveness and how to effectively leverage investments in web analytics technology. We were joined by, Steve O’Brien, Akin Arikan, and Karen Hudgins from Unica which sponsored the get together. What struck me most about our conversations (on camera and off) is how committed each of us is to moving Web analytics beyond what today is largely a passive, report-centric discipline towards one that:

  • Improves the effectiveness of individual customer interactions
  • Actively contributes to the understanding of the customer
  • Is a key part of improving marketing’s ability to measure across channels

Five key stages of web analytics maturity

One of the bits of original thinking that I contributed to the discussion was a framework that breaks down five stages of Web Analytics maturity.

Web Analytics Maturity Framework

Don’t pay attention to the stages for the moment… This is not a new picture (I pulled the base graphic from a from a client deck I presented in 1998). And, few would disagree. If a marketer is not leveraging any data to drive marketing communications, just adding a little bit of filtering or segmentation will have a tremendous positive impact on results. But, at some point, our ability to continuously improve results through segmentation levels off. At that point, the way to get the next hockey stick impact on results is to use individual customer data. Definitely not new thinking, but I think it helps level set us that is what “1:1” or “customer-centric” (you pick the cliché) marketing is all about – using knowledge about the individual customer to drive interactions that, at the end of the day, benefit both parties.

This framework can also help us think about how we are using the mountains of web data that we’re collecting to help us move up and right on the chart. I break the role and the progression of web analytics down into five stages:

  • Stage 1 – Site analysis: When we get started, we’re really just trying to get our arms around the data and the traffic on our site. The focus is to understand how visitors are getting to the site and what they’re doing there. But you also need understand why they’re there and whether they were able to accomplish what they set out to do. How do you do that? Avinash suggests that’s quite simple, ask.
  • Stage 2 – Site optimization: The goal, of course, is to avoid analysis paralysis and look for ways to leverage the insight we are gaining about how visitors access and use the site to drive more visitors to the site, to optimize the experience of visitors once they are on the site, and to help more visitors accomplish what they were trying to do.
  • Stage 3 – Segment targeting: As we continue to focus on improving customer experience, we inevitably start to look for ways to segment visitors into different groups either through data explicitly provided by the visitor or through insight inferred from the session and prior interaction data. We then apply the segmentation to customize visit experiences and target content.
  • Stage 4 – Individual customization: At some point, our ability to continually apply finer segmentation and impact results levels off. That’s when we start to apply individual-level web interaction data to customize online interactions.
  • Stage 5 – Integrated marketing: Of course, the holy grail of all of this is fully integrated and customer-centric marketing in which we seek to integrate insight from online behavior with what we know of an individual across other channels. And, we do this in order to inform and optimize all interactions – regardless of channel – with the individual.

The sad part of all of this is that few companies have matured their Web analytics capabilities beyond Stage 3. In fact, I’d estimate that 80% (not based on a quantitative study!) of firms are at Stage 1 or 2. Why? Well, it’s darn hard! There’s tons of data to wade through, the industry is learning as it goes, and the technologies that help marketers move up the curve are still pretty immature and poorly integrated.

So, how will this framework help?

Use the framework to understand where you are today and what you want to work towards and over what time frame. Each stage of maturity focuses on unique business objectives, requires a different level of analytical savvy, and demands different functional capabilities from your supporting marketing technology.

Over the next several weeks, I will continue to drill down on this topic with additional posts. Please add to the discussion by commenting and providing feedback on the blog or feel free to contact me directly.

Five steps to understanding customer retention December 4, 2007

Posted by Elana Anderson in Customer Analytics, Database Marketing, Marketing, Marketing Measurement, Marketing Strategy.
Tags: , ,
23 comments

 I recently responded to a question from a network that I participate in.

What is achievable customer retention and is there a level of customer retention that is not profitable to reach?

I’ve talked with a lot of marketers about this question and, frankly, there is no easy bake answer. It’s easy to look for a quick published statistic or benchmark and call it a day. But, how much does knowing that your retention rate is better than your competitor’s really help your business? It may help CYA, but it doesn’t help your bottom line.

IMO: marketers rely way too much on benchmarks (open rates, click rates, retention, etc.). Rather than rely on industry benchmarks (I don’t even know of a comprehensive source for retention by industry), I encourage marketers to:

  1. Establish a baseline for current average retention. Examine your customer base to understand average retention. Better yet, do it by customer segment if you can.
  2. Understand the timeline to customer profitability. Every business has different acquisition and services costs so if you don’t already know how long it takes for a new customer to become profitable, then you need to figure it out. Subtract your costs to acquire and serve the customer from average customer revenue over time. Companies that are really good at this use individual customer revenue and get into cost minutia to attribute costs at an individual level and even include costs like physical plant and electricity. But, if you’re just getting started, keep it simple and stick with averages.
  3. Set a target retention rate. The longer it takes to become profitable, the higher the retention rate needs to be. Establishing and monitoring a retention KPI will tie retention directly to business performance.
  4. Define marketing tactics to improve retention. If current retention is not at the target level, then set improving retention as a key business objective and drill down into a series of tactics aimed at moving the needle. Don’t shoot in the dark though. Engage a statistician to do some data analysis to better understand what key factors that correlate to longtime customers or customers that attrite. Then, establish marketing and customer service practices and campaigns that are specifically focus on encouraging the factors that are correlated with long-term customers.
  5. Measure results consistently. Periodically, reevaluate the retention rate to see how what you are doing is impacting customer retention. Make sure you are also considering metrics that help you tweak your programs at a tactical level too. Specifically, are the tactics you have implementing really encouraging those factors that correlate with long-term customers?

Home Depot: Say it ain’t so! November 5, 2007

Posted by Elana Anderson in Customer Analytics, Database Marketing, Marketing, Marketing Measurement, Marketing Strategy.
Tags: , , ,
add a comment

Home Depot Kids WorkshopIf you have kids, maybe you’ve heard of the free workshops that Home Depot runs the first Saturday of every month. My kids just love these events. When we go, they don their bright orange aprons (courtesy of Home Depot) and announce that they are going “to work.” This weekend, as the remnants of hurricane Noel hit our coastal town, my husband and I decided to head over. While he shopped, the kids and I built pirate ships (I only suffered a few hammer bruises). I also chatted a bit with the employee running the workshop who told me that corporate is planning to cancel the program. NOOOO!

Why are marketers forced to be so short sighted? 

Please keep in mind that I don’t have confirmation that the program is indeed being canceled – my intel here is a short exchange with an employee who may have been misinformed. But, it did get me thinking about marketing and how focused on the short-term marketers are forced to be. Sure, measuring the impact of a marketing program like this is hard. Maybe the short-term ROI of program is limited (although our exit bill was ~$120), but Home Depot execs shouldn’t discount the fact that the program fosters a bunch of little do-it-yourselfers who become Home Depot brand advocates at the tender age of 5.Before canceling this program to save costs, Home Depot needs to examine the impact:

  • According to Home Depot’s website, an average of 75 kids/store attend the workshops each month. With 2100 stores (and an average of 1.5 kids/adult), that’s 105,000 adults in the store on the 1st Saturday of the month that likely wouldn’t have been there otherwise. Now assume that 10% of those adults spend $100. That’s $1,050,000 each month. Now, consider the flip side, cancel the program and the parents don’t come in and spend – that’s an annual loss of $12.6 million.

  • It’s also not unreasonable in the least to think that the program has some additional brand impact on the parents of these kids. Again, let’s conservatively assume that 10% of the parents make one additional trip to Home Depot and spend $100. That’s another $1,050,000 per year.

  • Home Depot says that over 17.5 million projects have been completed in the workshops since 1997. I’m sure there are quite a bit of repeat visitors so let’s assume that the program has reached 6 million kids since it was initiated. That’s a lot of brand advocates with future purchasing potential. Once again, take a conservative assumption: let’s say 1% of those kids grows up to spend $300/year. That’s $18,000,000 per year (in today’s money).

Starts to add up, doesn’t it? Now, I don’t know much about Home Depot’s business or have any insight to average order size and that sort of thing so it’s hard to go on, but I think I’ve made my point. I wonder if the execs are evaluating the program based on costs and fluffy returns or quantifying the fact that the program:

  • Drives traffic into the store.

  • Yields (by my back of the napkin estimate) $13.7 million/year now.

  • Creates future buyers.

I sure hope the marketing folks at Home Depot are presenting the powers that be with numbers and not just shrugging their shoulders about the value of the program. Even though some of the numbers may not be auditable based on available data, I’ve always found that presenting numbers based on realistic (i.e., totally believable estimates) is a very powerful tool for getting senior executives on board with a marketing strategy or a program idea.