What I Learned from GiveSafe’s 3rd Lean Startup Experiment

Transporting an intern to one of our on-foot cold calls

My name is Jonathan Kumar, the individuals who helped me run this experiment are John Sechrest, Josh Schukman, and Adam Berk. The link to my original hypothesis is: https://trello.com/c/pKOzCKQx/105-my-name-is-jon-kumar-jkooms-i-saw-this-via-impact-hub.

I am part of an experimental incubator that gives startups and entrepreneurs funding in exchange for testing and sharing lean startup hypotheses.

I THOUGHT that 10 out of 50 SMBs based in downtown Seattle would donate time or money to seed our homeless app among their employees— ACTUALLY 3 of 50 SMBs did this after we contacted them over the course of 30 days (3 followups total).

The MOST SURPRISING thing I saw was the lack of response. Given the “local state of emergency” the Mayor declared recently about homelessness, and how sweet our project is, I thought more than 3/50 would respond within a month. I BELIEVE this happened due to executional errors on our part, but also a combination of it being cold outreach + people caring less than I thought. RESPONSE RATE needs to improve.

Lessons / Frustrations:

• Before the 30-day timer started, I should’ve prepared a master list of the companies to reach out to instead of leaving it open/semi-guided. I ended up doing this, but only with 10 days left in the original experiment window.
• Before the 30-day timer started, we should’ve prepared a list of strategies/email templates to use for initial outreach, cold followup, etc. I ended up doing this, but only with 4 days left in the original experiment window.
• Within the 30-day window, we didn’t learn what the best call-to-action was (we tested 3 different CTAs) because of so few responses. We also didn’t do enough followups to see how many were required on average to drive a response.
• We could have explored marketing automation tools, OR we NEEDED to block out more time to go through the no-response followups. Perhaps it should’ve been 7 instead of 3, or mix phone/in-person with emails.
• On-foot approach led to a warm response but few real results in short time.
• Learned that admin assistants can actually be instrumental in the process, and how working through admin associations, associations between buildings, or other professional groups may well be more effective than cold email or on-foot approaches.
• Warm outreach is WAY better, but obviously required the relationship built upfront. This experiment did not include companies I already know (for which I have >50% response rate when I reach out). We must either invest more time in relationship-building, or use that time to be better about cold followups.
• I picked a very experienced, talented person to execute the communications. With that though, came some expectations that I did not meet (see first two points above), differences in opinion on the emails we should send, and his busyness with other better-paid projects to be able to wholly focus on just this experiment. His time is quite valuable and he couldn’t afford to iterate / fly by the seat of his pants, as is my style. If I wanted to keep that, a younger candidate might’ve been more appropriate.
• UPDATE — Fiodor Tonti out of NYC suggested this as a learning: “The most important conclusion is when you have a small sample of leads, a qualitative approach is more efficient than a quantitative one. With just 50 leads, the time spent on crafting the perfect written communication might not be worth the ROI, given the already-meager baseline of email open rates for ANY communication (let alone a call for donation). In-person or cold-calling allows you to measure not only how often you’re failing but also WHY, and let’s you adjust as many times as the number of leads.”
  (I think the suggestion of considering the ROI on crafting emails on such a small sample is an astute one — and indeed, we did pour a lot of time into those with minimal results. Building towards ambassadors versus converted leads seems to make more sense as well. Thanks Fiodor!)

Outcomes:

Out of 50 SMBs (three emails sent to each), we attained 3 partnerships. 2 SMBs will donate time/money and 1 SMB will do an employee blogpost.

Next Steps:

We’re in the middle of a pilot launch, but I’d like to draw up and execute another lean experiment soon.