In my podcast yesterday, I had two Agile Development experts from Xerox. Several of the key points that they made was that Agile even though perceptively may be that it is undisciplined, it actually is a methodology that is driven by metrics and therefore very disciplined. They discussed how Design for Lean Six Sigma is applied within the development team to obtain the customer information needed.
This brought my thought process back to the outline by Eric Reiss where he believes that you need both a programming and customer development team. Though I did not ask the question in the podcast directly, I believe that Eric’s explanation and the Xerox definition closely resemble each other.
The important of gathering the correct metrics and the voice of the customer seems so important in the agile method. Since, I have been testing and applying Agile methods to the marketing process I studied Eric’s diagram in the terms of a marketing funnel seeing his cycle take place as we discover, validate, create the offer and then scale the offer. Taking this approach, I can see developing a marketing campaign around this. For an easy example, I will use a simple direct mail piece.
- Customer Discovery: You must decide who the customer is relative to the offering ; A simple procedure of target marketing or building a customer persona.
- Customer Validation: If you are planning a 20,000 piece mailing, than you would of course get an adequate sample size and to validate the offering and see if your call to action is strong enough. You may adjust the mailing several times to improve it and see which mail piece performs the best.
- Customer Creation: I would propose that you set your demographics tighter and determine the needs stronger for the offering. Maybe, a Mafia offering of sort, something they can’t refuse.
- Scale Company: You should have acquired enough knowledge to scale your offering not only to meet the needs of the customer but the resources of your organization. Maybe, it is something as simple as the final number of pieces that will be mailed and/or the material that you have on hand based on the response rate that you have tested.
Nothing earth shattering in this proposal, but the true iterative process will or should change the way you look at this. As Jim Highsmith, a founding member of the Agile Alliance and celebrated author once said: “Deliver the product needed at the end, not the one requested at the beginning.” The purpose of such a relatively small project is not completing the project but the ability to learn how to adapt a process to your customers needs. As you take these relatively small steps they will turn in to large leaps as time goes on.
Usually the biggest argument that people have against a process such as this, is time. However, my argument would be that there is a huge tradeoff between time and effectiveness. Worst case scenario is If your sampling never works and you end up sampling all 20,000 what have you lost, a little time? Another scenario is that you just mailed 20,000 and were happy with a typical return of 1 or 2% because that is what direct mail gets. So if you receive 200 to 400 responses you are entirely happy. But what if you sampled 1,000 and then another 2,000 and another 2,000? If you received 1% just on these and 3 to 5% on the balance, you would have increased your effectiveness from 200/400 to 500/800.
This simple concept should actually become a foundation of your mail system. Presently, I have started to formalize a plan for my own e-mail delivery system and breaking my list into sample sizes that I rotate for feedback. This way I will be delivering the same e-mail practically every day with an adjustment to each one based on the click stream from the day before. It may be nothing more than moving one section to a higher level in the letter. The next day, maybe I would add a picture to the worst performing section and so on. What do you think I will learn? Or not learn?