Making better use of online data

The data meeting
It's 9:24am Monday. You and your data analyst toiled all weekend to create a presentation with 60 pages of SEM/SEO/click-through/website stats/Facebook/BBS/seeding/email/widget reports to present to the managing director at 9am.
 
You had a mountain of information that you believed were equivalent to insights. You presented the data according to rehearsal until, on slide #12, your MD interrupted you with a simple question: "So what do we do with all this data?"

The meeting ends 36 minutes and 48 slides early.

Have a plan
Online analytics is a large and complex topic, and growing more so with the constant introduction of new technologies and digital marketing opportunities.

I often hear marketers complain about the lack of reliable data and good measurement tools as reasons for the data not being useful. While that is true to a certain extent, I believe the more pertinent problem is that many online marketers haven't spent enough time connecting data collection efforts with overall business strategy and execution plans.

One way to tackle this is to start with the end action in mind. Ask yourself, and your sales and customer service colleagues, this question first: "what actions do we need to take in the next few weeks/months to drive business?" Once you have a list of pre-defined actions and execution mechanisms in place, you then just need to gather the relevant data from online and offline sources to support your decision for executing the different scenarios. 

Automate it
Data has a limited shelf life, so marketers also need to act on the insights quickly while it is still relevant. A customer may be looking at your company's products today with the intent of making a purchase within the week. It's too late if you need a week to analyze the data, and another week to deploy a campaign to re-engage this customer.
 
One way to tackle this timing problem is to automate decisions and actions for your online business as much as possible. For example, Amazon.com's related content/product recommendations list was created with the intent of increasing the size of the average purchase on the spot, instead of cross-selling and up-selling later. You can create a simpler system, but the thinking process is the same: create pre-defined, behavior-based customer segments with different actions for each. Work this out ahead of time with your colleagues from different departments responsible for executing, using a decision tree and/or flow chart. 

What problems can you solve for customers on-the-spot so they don't have to come back to your website again or call your customer service center? What information can you readily provide them to save them time?  What sales and customer service process can you automate and put online without compromising customer satisfaction?

Better understand your customer segments
After you map out the courses of action, you can then improve your understanding of segments. What automated services do they need from your brand? Who belongs in these segments and what drives these people?  Behavioral-targeting information (e.g. viewed a product/typed in a search keyword) is more accurate and more readily accessible than demographic targeting (female, age 28) online, as it demonstrates needs at specific times.

The introduction of Web 2.0 technologies creates more opportunities for better understanding your segments' memberships and their relationships with other customers. As users participate in communities that you help to create, they give you a basis by which to measure their relative influence -- How much content do they contribute? How many people follow them and are in their network (e.g. on Kaixin)? The people who contribute often and have extensive networks are the coveted influencers that your brand needs to target and keep happy. Put them in a unique segment with an action plan different from those of other users.

The availability of near real-time data captured via Web 2.0 tools also allows you to better understand your customers' latest state of mind and desires. Through data mining you can further refine your company's segmentation marketing strategy (e.g. by targeting the keywords and topics that a customer often discusses with others). In the Web 2.0 world, the relationship a user has with others will show a lot more information than just what the user is looking at alone. By identifying key relationships, you can target and motivate not just one person, but potentially entire groups of people with a strong affinity for your company's product.

The above is not an exhaustive list, but it will ensure that you focus resources to gather only the data you need to take relevant actions. Once you have this information, you also know exactly what triggers you can pull, by segment, to drive business online or offline. This will minimize the risk of your company getting lost in a sea of information without knowing what to do next.

Both your analyst and MD will be pleased with a much shorter and focused presentation next time.

Andy Chang is managing consultant at Precision Digital Marketing.

 

Comments

Kristi Barrow
Kristi Barrow April 22, 2009 at 7:29 PM

Andy,

excellent post! You were spot on with the lack of "connecting data collection efforts with overall business strategy and execution plans."

I also really believe that it is a lack of resource issue in a lot of cases. Marketing analytics resources are not given the priority in most companies, yet it can be one of the most cost effective ways to increase the bootom line. I think there are a lot of marketing managers running around trying to just produce reports, so sadly they have no time to analyse and develop actions based on the data.