Rauch Digital - Hypothesis Driven Marketing
Michael Rauch

Michael Rauch

Share on facebook
Share on twitter
Share on linkedin
Share on email
Share on print

Hypothesis Driven Marketing

Hypothesis Driven Marketing takes its roots from Agile practices, The Lean Startup, Design Thinking & Growth Hacking creating a way for teams to test & learn toward growth for their product or service.

How do we ever truly know if our marketing effort hits our target? Through my experiences the only real way is to just get going! Dare I say…hypothesis driven marketing?

With that said, applying some principles around how to execute this learning can help teams. This is NOT meant to be a process to follow. Rather the intent is to offer a framework for teams to establish some ground rules for better collaboration & focus.

First we need to start by understanding some of the frameworks, philosophies and ideas that have been brought together in this example.

The roots of Hypothesis Driven Marketing

Agile principles

The framework of Agile was born back in 2001 by a group of 17 people met to talk, ski, relax, & try to find common group. The were looking to develop an alternative to the documentation driven, heavyweight software development processes. This group produced four elements to create the Agile Manifesto, which has been used by teams all over the globe.

Many processes were represented at this gathering. There is Scrum, Kanban, Lean, xTreme Programming (xP), and many others. All have their place, but none of these fully fit a hypothesis driven marketing model.

However, when we look at the core principles to apply it really comes down to the following three key concepts. Focus on customer value, validated learning & transparency.

The Lean Startup

Another widely adopted practice by product development teams is the concept of The Lean Startup, a book written by Eric Ries in 2011. Ries outlined his concept for focusing on the customer. Through delivery of minimally viable products (MVP) to the customer, product teams continuously shift their strategy by pivoting based on validated learning. This is done by continuously engaging with the customer.

Applying those principles to hypothesis driven marketing, teams should understand their north star (or mission), and then create a strategy (hypothesis) on how to achieve these results. Then the team learns through executing a build, measure, learn feedback loop. Based on these learnings, continuously adjust the strategy to best deliver value.

Design Thinking

The background of Design Thinking has taken many forms. Some origins date it all the way back to pre-1960. Conceptually teams must focus on the customer first and use the customer as the backstop for decision making.

However, when we think about hypothesis driven marketing following the concept of working through the cycle of empathize, design, ideate, prototype & test makes sense. Truthfully, this concept is likely the easiest way to begin applying hypothesis driven marketing.

Growth Hacking

Growth hacking was first coined by Sean Ellis in 2010. The idea is for the team to first ensure that true product market fit is understood. Also that an efficient conversion process has been established. From there the team needs a role that’s true north is growth. This role focuses on scrutinizing everything they do by understanding the potential impact on scalable growth.

This role is entrepreneurial by spirit and throws out many of the preconceived ideas about how marketing has been done in the past. Through prioritizing, testing & analyzing ideas this role finds scalable & repeatable ways to grow a business.

Structuring hypothesis driven marketing

So how can we take all these similar but differing concepts and meld them into a way to organize marketing of a new product or service? There is a lot of talk about the customer. Many concepts focus on short cycles that create opportunities to test and learn. But how do we start?

Creating your hypothesis

First we begin by taking ideas from brainstorming, customer engagement, and past experiments and create a hypothesis. One concept from a blog post by Joshua Partogi on Scrum.org is the write backlog items as a hypothesis. This allows the team to focus on more than the possible solution.

Whether it be on a sticky note, spreadsheet, or my preferred tool Trello, start by documenting the hypothesis driven backlog items in this form:

We currently believe that (objective) will be achieved if we deliver (possible solution) because our customers will gain (customer value). We know we’ve achieved this when (consumer outcome) and (value driven metric).

Categories

The next step is not a necessity, however I find it a useful element when the backlog grows rather large. The team can collaborate on categories that each backlog item falls into. This will help visually identify if the team is focusing too much or too little on elements. Here’s a quick list of some potential categories for hypothesis driven marketing for growth of a web application.

  • Acquisition
  • Engagement
  • Retention
  • UI/UX
  • Performance
  • Brand

Now each of these categories may or may not apply depending on the product or service you are attempting to grow. Use these as a starting point, and adjust for the needs that fit your team.

Estimate Sizing

Another valuable element comes from Kanban principles. With Kanban teams want to strive to limit the work in progress. This keeps teams focused and working at a sustainable pace. To do so we need something to measure and size our hypothesis driven marketing statements. However, creating hours based estimates for a hypothesis can be difficult and likely results in a pure guess.

With this in mind, create estimates based on relative sizing against all other hypotheses that exist in your backlog. As the team does this use a simple scale, for example “t-shirt” sizing.

  • Small
  • Medium
  • Large
  • X-Large
  • JUMBO

Create a list of the tactics to use

A typical hypothesis may be slightly abstract. It may not jump off the page as to how the team will go about proving this out. Likewise it may take multiple divergent efforts to achieve your learning results. That is where the tactics come in.

Creating a simple list, describing what will need to happen to prove things out is all that is needed. Again taking an example from a web based hypothesis some typical tactics may look like this (albeit slightly generic for illustrative purposes here):

  • Run Google AdWords campaign
  • Create A/B test for the product page
  • Develop back-linking strategy for SEO
  • Create landing page to drive traffic toward
  • Conduct consumer research through usage survey

Attempt to prove your hypothesis

Now this is the fun part! Put your plan into action. Sometimes this is in the form of surveys. Maybe is it in the form of launching a new feature to a website. However, the key here is to make it measurable. Ensure that data can be drawn out, analyzed and acted upon.

As we think about web based hypothesis driven marketing, tools such as A/B testing, heat mapping, and link management provide actionable results. Some of my favorites include CrazyEgg, Optimizely, & Bitly.

Other options may be just as simple as crunching data you have been collecting all along. Google Analytics, Google WebMaster Tools, Moz, & Screaming Frog are all tools that can offer insights into your existing web configuration and traffic.

Many other options exist such as SurveyMonkey or detailed analysis of your emails deliveries through MailChimp. Even looking at abandoned carts in your eCommerce store and attempting to convert those potential customers through personalized reminder emails.

It’s either a Win or a Learning

After we execute each hypothesis we track them as either a win or a learning. Wins are something that we tried and hit our goals. Teams often take these wins and use them to create more wins. Or potentially just create a new hypothesis that simply adjusts the measurement metric to further grow.

On the other hand. Learnings are things which maybe didn’t hit our intended goal but we learned from. Turn these into a different hypothesis. Or maybe scrap them all together. The point is, these are not a failure in the traditional sense of the word. The old quote goes:

“I haven’t failed — I’ve just found 10,000 that won’t work.”

– Thomas Edison

Where do we go from here?

It is pretty simple really…get started! Embracing a culture where “failure” is valued as learning rather than in a negative light is really that secret sauce.

At the time of writing this post, Rauch Digital is engaged in a couple projects applying these principles. Check back for future posts breaking down the results. Until then, reach out if you are thinking about the best way to execute hypothesis driven marketing.

Share this post

Share on facebook
Share on twitter
Share on linkedin
Share on print
Share on email

About the Author

Michael Rauch

Scrum Master by day, digital marketer by night.  My hope is that I can share my experiences and give readers a different perspective on the topics I cover.  I am constantly learning and trying to improve, so come along for the journey!

Rauch Digital Marketing - About Page