Andrus Purde


Can you validate a startup idea “by the book”? Here’s how Outfunnel got started

It’s September 2017. I’ve just quit my rather exciting job as head of marketing for Pipedrive, a sales CRM. I have a rather vague idea of a marketing tool that should exist but I don’t want to waste my own time or that of others. (I had wasted enough time on 4–5 wantrepreneurship projects over the years). I was prepared to fail, but if I was to fail, I wanted this to be for a better reason than “oops, it seems that the product we have built is not needed”.

I knew I needed to validate the idea. Which meant I needed to speak to a large number relevant people and I needed to have a good system for getting this done. Here’s what happened.

How I sourced the list of people to talk to

I wanted to speak to people across multiple job functions because I wasn’t sure which precise problem to tackle. And the problem needed to be precise because there are 7000 marketing tools out there already.

I also wanted to speak to people in multiple countries, and ideally a good mix of phone and in-person interviews to get a better sense of who I’d be targeting.

At the time I had 1700 or so LinkedIn connections, and a quick glance showed me A. there were people there that could be in the target audience B. I didn’t really know most of them.

So I downloaded my LinkedIn first-degree contacts as a spreadsheet. I was looking for CEOs, entrepreneurs, sales andmarketing leaders, and business consultants. Browsing through the list left me with about 120 contacts, which I imported into my CRM tool. From there, it was dead easy to start new conversations since I already had all the contact details and my notes from browsing the list handily available.

(Please note that LinkedIn stopped including email addresses as an included field as of November 2018 which is both good and bad news).

Second, I asked for intros whenever I remembered and dared to do it. Often the response was a polite “not sure,” but overall, I’d say the old axiom “no harm in asking” stands, independent of whether you’re chatting in person, on the phone or via email. A couple of people shared my ask with a marketing or entrepreneurial group they belonged to, and this got me 5+ solid leads.

Third, I had some speaking gigs lined up and I knew I would meet some brand new people this way. So I set up the contact capture form with some automation that I’ll describe below.

Fourth, I hustled some contacts from relevant conversations I picked up from social media and a relevant Facebook group.

Fifth, I also drove some people to the beta landing page with Google Ads.

How I automated the workflow, using Pipedrive, Calendly, Zapier and other tools

I started by collating all my interview prospects into Pipedrive because I wanted all the information to be in one place and I wanted to be able to track my progress.

When I emailed people, I almost always added a Calendly link, so those interested in talking to me could pick a suitable slot without the back and worth. And voila! interviews started appearing in my calendar as long as I stuck to my goal of doing 10+ first approaches per day. I followed up at least once with everyone I hadn’t heard back from, and sent up to five emails to people I really wanted to speak to.

To capture inbound interest from speaking gigs and other sources, I set up a simple landing page with Voog and added a Paperform signup form. When people signed up, their data was automatically sent to Drip so they would receive a welcome email, and a follow-up email. If they opened either of these emails, their data was sent to the Idea stage of my customer development pipeline. Then I could reach out to those who looked even semi-serious about improving their funnel game.

Asking qualifying questions in the signup form proved useful. I asked about their main marketing challenge(s) and the marketing tools they were using, and both helped to later understand which prospects were worth chasing and who was “just looking.”

How I conducted interviews

I was learning from day one, but I also felt I wasn’t making the best possible use of time in some of my conversations. Then I was recommended “The Mom Test”,a beautiful, concise, no-fluff book about doing customer interviews. I had also read the “The Four Steps to the Epiphany” but the latter is a drier read than most tax filings, so I doubly enjoyed “The Mom Test”.

If I could take a stab at summarising “The Mom Test” in one sentence, it would be “Don’t believe what they say, only pay attention to what they have done and are doing.” There, I’ve saved you 10 bucks and a couple of hours of reading time. (But I’d still recommend to get the book).

A meta-learning I’d like to share is that doing 100+ conversations in a short span would have been a waste of time if I just took notes and didn’t create structured data out of those notes.

I typed my notes into Pipedrive after each call or meeting, or at the end of the day when memories were still fresh. I also set up some custom fields in Pipedrive so I could have helicopter-view insights later. More specifically, I aimed to fill the following fields for each interview:

  • Main challenge, or core business needs as expressed by the person (or sometimes inferred by yours truly)
  • Surprising insights. Anything unexpected I learned about competitors, possible future features or services
  • Needs. Features (of my planned product) that could be useful for that company

I also captured data on tools used in sales and marketing, industries, and lead sources.

About 30 interviews in, I hit a wall

Things started out good. Not rosy but good. I was out of my comfort zone and doing cold approaches. I had started doing interviews and having conversations, some of them with really smart, interesting people.

I was asking useful questions and getting insightful answers. I was making notes and creating structured data.

The slight problem was: I wasn’t getting closer to learning what to build and whether I should build something at all or just get a job.

The insights I was collecting were all over the place. The people I was talking to had vastly different funnels, objectives, and concerns. There were very few common threads. After about 30 conversations I felt I should build 30 different products.

Doubts started lurking in the corners of my mind. Maybe I’m not good at that customer development thing? Or that entrepreneurship thing..

I remembered a customer research axiom that you should continue doing interviews until the answers you get start repeating, so I kept going. Luckily, answers soon started to repeat around a couple of relevant themes. By about interview number 70 I had a pretty good idea of what to build.

So I conducted 100 interviews. What happened next?

First, let’s look at some numbers, which is easy if you use a CRM tool such as Pipedrive. As you can see I’d collated a list of 317 people to contact, and I got in touch with around two-thirds of them. About half of them responded and were available for a chat. Out of these respondents, 58 people had unmet needs that could have been resolved by one of the product features I had in mind for Outfunnel.

Jumping ahead, I got back to 21 of those people later with an offer to try our early beta, but only one of them has started a paying subscription to date. I guess this proves my earlier point that what people do is very different from what they say.

As you can see, I got started in November of last year and got to full speed in December, but then slowed down as Christmas holidays neared. (More the case of people not wanting to talk to me than the other way around).

January was a full on interview month, my record was 10 scheduled interviews per day. My earlier noted slump also appeared in January, and disappeared a few weeks later. By early February, I had a pretty good idea of what to build, and while I shifted focus to finding people to help me build a prototype, I continued doing some outreach and interviews in February and March.

We began building In April, and I started re-approaching some people with an offer to try our beta, which we launched in May.

In August we launched our web tracking feature and I haven’t stopped doing customer discussions while we’re building the rest of Outfunnel, a sales-centric marketing tool.

A happy ending?

The customer development process and this post had a happy ending. I validated the problem, some ideas for solutions and we built and launched a prototype. We now have 50+ paying customers using Outfunnel, and we’ve raised a seed round from top investors so we can keep ploughing ahead.

The thing is, this is far from a happy or any other type of ending for Outfunnel. It’s great to have launched something and to see dozens of companies using the product, but the end of the process described in this post is just the beginning of Outfunnel — so wish me luck.

I hope you found this story useful. If you’d like to contribute to shaping the rest of Outfunnel’s story, please get in touch.

PS. We’re hiring! We currently have five people involved with Outfunnel in different ways, and we’re actively looking for early employees and even late co-founders. In the next six months we need to hire front-end developers, full stack engineers, product managers, designers, marketers, and customer success people. Ideally people who like marketing and marketing automation. (I guess that may rule out us ever hiring sysops …)

This post was based on my recent talk at SaaStock and an earlier post on my personal blog

Photo by Alex Iby on Unsplash

More by Andrus Purde

Topics of interest

More Related Stories