"Computers are useless. They can only give answers."

That famous quote comes from an unlikely source: artist Pablo Picasso. If you think statistics is a subject for the numbers geeks, you're missing the point. It's about so much more than that. Back in the days when computer storage space was precious and costly, the phrase "fact-based decision-making" hadn't been born yet. Managers were prized for their "golden gut," or the ability to make decisions with good outcomes in spite of a lack of information. Today most companies understand the value of managers with superior quantitative abilities, coupled with leadership, organization, communication and other key skills. The age of "quant" is responsible for the proliferation of MBA's, as well as job notices that declare, "MBA preferred" as a qualification. While those who understand the more subjective art of decision-making may decry a quantitative approach as mechanical and limited, it's hard to argue with the facts.

Plan
In this age of technology, you might assume that decision-making has been largely relegated to computers, but this is not the case. As Picasso pointed out, computers "can only give answers." While this sounds like the whole enchilada of making good decisions (if you have the answers, what else do you need?) it's not even close. The Plan stage incorporates understanding the question, first and foremost. That's what draws the line between an entry-level report jockey and a rock-star analyst (and by analyst here, I'm including managers who are using analysis). One person answers questions; the other asks. If you have access to the right data, answering questions is the easy part. Knowing what questions to ask is what defines the superior business person and it is what defines excellent management skills. This is not the part of the equation that can be "farmed out." Neither the computer nor a technical whiz kid can serve as replacement for the experience, judgement, and business acumen required to ask the right questions. 

Do
If you are using experimental design to create advertising campaigns or using a stats package to predict returns on financial products or leveraging web analysis software to optimize your website's ROI, you understand the complexity and technical know-how required to turn data into metrics. But you also know this isn't the end game. Just search on the term "web analytics" and you will find article after article about companies that have reams of metrics and data that they don't know what to do with. The actual "doing" of statistics is the easy part, relatively speaking.

Report
Interpreting the results of any analysis is an area fraught with danger. This is where the best planned analytic project can fall apart. Drawing inaccurate or illogical conclusions will yield disastrous results, from wasted effort and investment to a failed business model. Statistics are like nature: you can believe whatever you want to, but it doesn't change the truth. Knowing how to take analysis and turn it into good decisions is far more critical than selecting the right software package. Yet companies will scrimp, nickel and dime on employees and their training, all the while launching six month committees and appropriating seven figure budgets to buy the right web analytics/business intelligence/statistics/[insert your company's pet project here] software.

Act
You might believe that execution made as a result of analysis is the last step in a linear process, but here again is a place where many companies lose. Having a cyclical process with measurement plans built into the implementation phase creates an environment where a company can make decisions an organic part of the process and, as a result, remain competitive and in touch with the customer base. Payoff: better ROI, loyalty, market leadership, etc.

So, while you may be able to hire great analysts and buy the latest great technical solution for the "doing" part, there's no substitution or shortcut for talented, experienced, bright decision-makers who know what questions to ask, what metrics to track, how to interpret analysis and who are absolutely committed to implementing plans that include systems for measuring their effectiveness along the way. 

Lather, rinse, repeat. Forever.

Loading mentions Retweet
Filed under  //  analysis   analytics   decisions   management   MBA   quantitative   ROI   statistics   website  
Posted 6 months ago

Social media isn't a fad. It's just media and customer karma.

It's funny to read all the speculation about the destiny of MySpace, Twitter, Digg, etc. It makes you picture boardrooms across America, nay, the globe, discussing whether their companies should start a Twitter account. Almost as plentiful are the articles giving advice about how companies should be dealing with social media, what it takes to be a "social media expert" (as if such a thing existed), or whether social media is replacing search.

The bottom line is this: there is one simple truth. The rules for social media are just like the rules for every other kind of media.

1. You decide what media your company should be in based on what media your target markets are in.

2. Your target markets and resulting media strategy will reflect the personality of your brand.

3. If you're testing your CPA (cost per action or cost per acquisition) across all media, you'll know when it's time to invest more in a particular media, no matter what it is. Whether the expense is a media buy or a staff to engage with your customers via social channels, you can still calculate an ROI. If you want to control the ROI, make sure your staff time-boxes their activities in that channel. 

It really is as simple as that. The questions are no different now and they won't be any different when the next big thing comes along. I don't mean to come off like a crazed MBA and make it sound like numbers are the only thing that matters. There is something else. It's called customer karma, and that's where the exception comes in. The exception is in customer service and managing your brand image. The first part of which you are compelled to do in social channels regardless of that channel's marketing ROI.

And the second part of which is simply not possible. "Your brand is whatever your customers say it is."(1) So whatever else you do, you'd better get out there and listen, not just to what your customers are saying about you, but to what they are saying about your competitors, too.

(1) Marketing in the Groundswell by Charlene Li & Josh Bernoff

Loading mentions Retweet
Filed under  //  CPA   customer service   marketing   media   ROI   target market  
Posted 7 months ago

We need a new word for "optimization"

Many would agree: SEO's are the rockstars of the web marketing world. [Post editing note: This applies to good, real SEO's. Not the snake-oil salesmen who try to pass themselves off as SEO's.] Not that they shouldn't be. Ethical SEO's provide a critical service to website owners and web surfers alike, when they help websites find targeted visitors and make it easier for visitors to find exactly what they're looking for. But over and over you hear, "Internet marketing is a 'numbers game'." What does that mean? To a certain extent, all marketing is a numbers game. But a weak way to play the game is to get as many people as possible to your site in the hopes that you will get lucky and somebody will convert. That's just not very smart, because it can be extremely costly, thereby driving down your ROI and wasting the time of those who don't find your website of interest.
 
Enter "the other kind of optimization": website optimization. If you are looking for more information on website optimization, good luck. You will be swamped with information about how to improve traffic to your website with search engine optimization. Is website optimization the redheaded stepchild of the web world?
 
It's a mathematical fact that it is cheaper to have a higher conversion rate than it is to get more traffic. SEO's are important here, too. They are the ones who get you the targeted traffic that has a higher interest level in your product or service. From there, the types of things you can practically test are somewhat dependent on your audience size. If you are Amazon, you may test every dot and tiddle; You have the numbers to deliver a statistically valid result from which to make a decision. If you're a little smaller than than that (most websites are) then you have to be a bit more judicious about what you are testing. Since you may not have the opportunity to dynamically generate an experience based on a visitor's past history, you will probably be focused on figuring out what motivates your various target segments to take a desired action. Both disciplines require the cooperation of your creative team, copywriters, web analysts, IT, etc. Both disciplines require careful planning and consideration of the overall marketing strategy. But they NEED each other, otherwise the SEO's beautiful traffic gets wasted as it heads out the back door and the website optimizer's tests never get enough numbers to be conclusive.
 
Everyone involved in marketing on the web has heard of SEO by now, or, the less frequently used "search engine marketing." I think we should find another term for website optimization, because it's confusing to many people. Any suggestions?

Loading mentions Retweet
Filed under  //  marketing   ROI   search engine optimization   SEO   web analytics   website optimization  
Posted 7 months ago

Marketing Execs Struggle to Show ROI

Really? Data availability and integrity is the number one barrier to showing ROI? Technology and infrastructure is number two? This is shocking! When so much great information is available, when web analytics and website optimization tools are FREE, this is very hard to believe. Is it possible that the survey results are dead wrong? That the issue is a lack of people who understand the data or know what to do with it?

One big gap that companies struggle with is inter-channel results, meaning, someone goes online to research a product, then finalizes the purchase in the bricks-and-mortar store. But hasn't it always been the case that retail purchases could not typically be attributed to a particular broad-market advertising activity? That may be a part of the ROI of the site, but you have to get creative about what you are capturing. You may not always be able to capture the sale itself, but if you can at least capture the fact that the visitor achieved their purpose on your site, that is one type of ROI that is also important.

Loading mentions Retweet
Filed under  //  data quality   Marketing   ROI  
Posted 7 months ago

Don't hire me to do your web analytics!

As much as I love wrapping my mind around the results of a well-executed, carefully measured marketing plan, analytics are only a means to an end. The point is to make smart decisions, period. Better decisions than your competitors, one would hope. You ask questions of your markets (prospects, loyal customers, even listeners) and they answer honestly, unfailingly, in the form of statistics. That is, if you do it right and if you know how to listen.

 If you're not making decisions with the metrics you track, you need to have a hard talk with yourself. Because if your metrics aren't actionable, there are a thousand other more important things you could be doing with your time, liked understanding your competition better, understanding your customers better, or better yet, understanding your competition's customers better!

 Here's one approach to help you get focused. Before looking at your web analytics reports, try coming up with business rules for the decisions you will make about each one.

   1) Open up a spreadsheet
  2) Down the left hand side, list each metric you track for your digital marketing activities (PPC campaigns, email campaigns, social media sites, etc.)
  3) Across the top, list columns: Business Goal, Result, Tolerance 1, Action 1, Tolerance 2, Action 2, Tolerance 3, Action 3.
  4) For each metric, list the business goal that metric addresses and how you will know when you have achieved it.
  5) For each metric, select appropriate tolerances and define the actions you will take if your results fall within these tolerances.

 If you think this takes too long, maybe you have too many metrics. If you are focusing on the critical few, this will be a quick and simple exercise that will help you focus your thinking and will make your report review go very quickly. If you have metrics that don't fit neatly within this template, this process should at least get you started thinking about how to make your metrics more actionable. You can take the same approach with each of your target segments. It may sound a bit mechanical, but it beats swimming in a sea of data and not knowing where to start.

 Start out with a stand-out product or service, add a genius marketing plan, execute and manage that plan with discipline. Every step along the way can be improved with great analytics.

Loading mentions Retweet
Filed under  //  consultant   marketing   marketing strategy   ROI   web analytics  
Posted 7 months ago