When Software Companies Think Like Agencies / by Dan Weingrod


One of the best parts of the Lean Startup sessions at SXSWi this year was not just the presence of the Lean All Stars, (Eric Reis, Dave McClure and Steve Blank), but also how true the sessions were to the tenets of Lean by presenting real case studies and prototype discussions. This ranged from very specific examples to expert panels and even to a King of the Apps showdown, (complete with crown and robe), focused on going beyond ethereal process discussion to real issues of failure, pivot, learning, iteration and success. The best of these case studies was Manuel Rosso’s “Concierge MVP” presentation, because it made me realize that for all the talk in the Ad world about agencies needing to think like software companies, software companies have grabbed this bull by the horns and are already thinking like agencies. Rosso is the founder of a web site called foodonthetable.com. Hs case study was about was his team’s unique approach to customer development and creation of an MVP.  An MVP, before you think this is a sports column, is a “minimum viable product”, or a product with the smallest feature set that will appeal to customers. What was unique about Rosso’s approach was how they achieved the MVP quickly, effectively without spending money on additional resource and by essentially using a variant of research techniques that agencies have been using for years.

Foodonthetable’s essential proposition is that weekly meal planning is difficult, time consuming and costly and that their target audience, Moms, needed a simple effective way to approach it. Their idea? Create a website that maps sale items at local supermarkets to a customer’s food preferences, relate these to a set of tested recipes and create a healthy meal plan and shopping list. In the digital world this value proposition might be the time to start coding a prototype. In the agency world we’d likely take this idea, do some broad market research and put it all in a brief to hand off to a creative team.  What Rosso and his team did was to start with one customer and with this:

They found a Mom in their target group who they thought would be interested and sat down with her in a Starbucks across the street from her local supermarket. Together they stepped through the above page and input information about where she shopped, what her food preferences were, how much she wanted to spend and other information. They had in hand the supermarket’s weekly sale circular and used their laptops to look up recipes online and a printer to print out the shopping list.  Everything was done by hand, no code was involved, no logos were created, (note the “Food Site” logo in the upper left corner), and no tag lines or copy written. Based on the scrawled writing on the form it’s also clear that there was plenty of room for improvisation and change based on actual customer feedback and physical realities. Do we need photos of the food? Let’s write that in for now. Should we add date and time? We can scrawl that in over here.

As time went by the form was adjusted based on customer feedback and actual experience until they added a second customer. The second customer added more variety to the feedback loop, but they continued to use this analog “concierge” approach to define and build a product through actual, real world, experience. Only when they got to the third or fourth customer, and an additional supermarket, did they realize they actually needed put something into code. In this case it became clear that having a database of local supermarket sales and specials would actually save them time over the analogue process of eyeballing weekly circulars in order to build menus.

The result was that after an eight week process Rosso and his team were ready to actually begin the process of coding and automating the processes that led to the finished site. The site itself is rapidly growing across the nation and features a home page one of the best examples of a clear, “why you should sign up” value proposition I’ve ever seen.

When progressive agencies and marketers think about integrating with “digital” or “technology” we often consider the benefits of using technology related processes already ingrained in tech companies. What the foodonthetable.com experience reveals is that we already have many of the right processes and thinking, it’s just that software companies just might be doing a better job of using them. As an ex-CPG marketer Rosso understood the value of consumer research, but he also knew that doing it in the traditional manner would take too long, be too costly and would not deliver the necessary opportunities to test and iterate. What he did understand, and what is at the core of customer development, was that the only way to succeed was for him to get out of the office, listen to customers and create a product based on small, fast iterations and feedback loops. And if it didn’t need to be polished or coded, so much the better.

Agencies have all the tools to implement these approaches, but our culture and traditions don’t seem to allow it. We listen to customers in focus groups instead of individually, we go through long, torturous research on customer needs and profiles without actually talking to customers. Worst of all, we then take our valuable research and present clients with a “finished” recommendation that calls for costly investments that offer no opportunity to revert or iterate. We might consider paying attention to the fact that software companies are using the tools that we’re traditionally expert at to make small bets, fail more quickly, spend less money, but ultimately end up with better results. Instead of worrying about how agencies need to think like software companies maybe we should think about how software companies can teach us to make better use of our own tools.