Data-driven startups: How to design and build smart funnels within your product

Will Little

The dance between product and growth in a new startup is rarely elegant, especially when founders fail to bake smart funnels into their product.

Ideally, the dance is best choreographed by a relentlessly pursued operations cycle:

1. Smartly defining funnel stages and Key Performance Indicators (KPIs)
2. Baking these measurements into your product
3. Implementing daily reporting mechanisms
4. Conducting tests
5. Responding to data by reprioritizing product and growth tasks accordingly

(…and repeat, ad infinitum.)

It's extremely common for new founders to either flat out ignore this process, or fall into bad habits by not being comprehensive and/or ambitious enough along the way.

So, let's dive into each step:

1. Smartly defining funnel stages and KPIs

Most digital products — really all products/services — have at least two funnels that business operators should pay attention to: customer acquisition and customer engagement.

A typical customer acquisition funnel will have stages such as:

• Total website page views
• Unique website visitors
• Email address acquired (e.g. on a lead-gen page or contact form)
• Signups
• Custom on-boarding stage 1 (e.g. fill out basic profile info)
• Custom on-boarding stage 2 (e.g. indicate interest in X)
• Initial dashboard visits
• Performs some action to be considered an “acquired customer” (e.g. creates something, shares something, buys something, etc…)

In addition, a typical customer engagement funnel will look like:

• Responds to a call-to-action in a newsletter, notification email, or other internal marketing campaign
• Visits website
• Signs in
• Performs key action 1
• Performs key action 2
• Performs key action 3
• Performs your most important action (e.g. upgrades to a paying tier)

Using the example app/company we've been talking about throughout this series (i.e. an app to help make meetings more efficient), customers will be considered “acquired” once they create a meeting and invite their first team member. That final stage is a great example of a KPI for us to track, along with the amount of time it took to acquire the customer after initially landing on your website (check out Google Analytics' User ID feature for an example how to get this data).

Also using our example app, for the customer engagement funnel we'll want to track sign-ins, meeting creations, users invited, and ultimately the stages of upgrading to a paid tier (e.g. selecting a tier, getting to the payment page, and processing the transaction successfully), AND successfully making payments month after month.

For this example business, the One Metric that Matters (OMTM) will be the number of paying customers per month. This KPI will guide all of our product decisions and experiments.

2. Baking these measurements into your product

Now that we have a solid first draft of our funnels and KPIs to track, now it's time to ensure your product team efficiently builds these measurements into your product.

While larger companies will use a massive Customer Relationship Management (CRM) product like Salesforce combined with something like Hubspot or Marketo to move customers through funnels, these services are — let's be honest — too expensive for the pre-funded startup.

Thus, in the early days, founders usually end up rolling with their own data (combined with Google's free Analytics product) and hope that they cover a sufficient amount of stages/KPIs to learn about their users and how product engagement is going.

What I'd recommend is, indeed, smartly design your funnels/KPIs and roll with your own custom, temporary solution (and study Google Analytics and Google Tag Manger carefully).

This is important mainly so you understand your business as effectively as possible and utilize free software out there.

As you get money in the bank, it's almost always better to move on to third-party software (e.g. I'll give a shout-out to Intercom as a solid option once you're ready to drop ~$100/mo on marketing + customer engagement software).

3. Implementing daily reporting mechanisms

It's surprisingly easy to go through all the trouble of carefully defining funnel stages/KPIs, bake them into your product, then fail to take the last step of actually ensuring the right people on your team can see the data and take action accordingly.

As a first step, build out your daily background tasks to collect the data into your database (e.g. working with the Google Analytics API is relatively straight-forward). Then, create a simple admin view that displays this data in a way that makes sense for your business.

What I usually do from there, then, is automate the process of dumping this view into an email on a daily basis and send it off to the right people on the team.

You can take advantage of email notifications from free/cheap software out there, but I've found it's better — if possible — to roll with your own (even as you start purchasing third-party software).

4. Conducting Tests

Now here's the critical part, once you have the data in front of you every day, what do you do with it? How do you decide what to work on next?

The answer comes back to how you've defined your OMTM and KPIs. The growth people on your team are going to be mainly focused on “the top” of your funnel (e.g. running experiments to increase traffic), and your product people are going to be focused mainly on stages closer to the bottom (i.e. within the product itself).

Where exactly you decide to focus efforts should depend on what the data looks like. If you have a weak spot in your funnel (e.g. users are having trouble performing some key action), then that's an obvious place to focus.

If your funnel is running smoothly, then it's time to run tests and pour gas on the top of the funnel and have your product people focus on the earliest stages possible.

A great common example is “A/B testing” different copy/images/button placements on your key views to see which ones perform better.

Or, closer to the top of the funnel, running ad campaigns across multiple channels (Radio, Facebook, Adwords) to see which ones effectively drive traffic and customer acquisition.

5. Responding to data by reprioritizing product and growth tasks accordingly

Finally, as you run tests, collect data, and compare funnels, make sure your product and marketing managers are reprioritizing tasks in their queues on a weekly (if not daily) basis.

A common point of operational friction in a company, especially as it grows, is the lack of coordinated information flow between product and marketing people. This results in chaos; tasks get completed that employees have enjoyed working on (or not), but have little relevance on the overall business objectives.

A smart idea is to be radically transparent about exactly what your funnels, funnel stages, and KPIs are at any given time so everyone in your organization is on the same page. This allows for individuals to make smart product and growth choices in the face of ambiguity or a lack of direct guidance. Setting up this kind of operational structure has a built-in buffer for management problems, which decreases friction and allows your business to run as smoothly as possible.

Author's note: this is the 12th post in a series of articles outlining a framework for pre-launch startup operations that my partners and I at Prota Ventures have developed. We're building a SaaS product on top of this framework to help founders execute new ventures and get connected to co-founders + advisors/investors along the way. Jump into the private beta queue here. Also, feel free to subscribe to my newsletter and I'll let you know when I get new content up. Thanks!