I tried to revolutionize travel with a Telegram group but ended up with a lesson in experiment design.
I was planning my 2023 New Year’s Hampi solo trip in November and realized there are no good community-based tools for solo travelers. So I decided to take matters into my own hands. A social tool for solo travelers seems like a simple enough idea.
A Telegram group for all the solo travelers visiting Hampi for the new year seemed like a decent enough MVP. The goals of the group were simple.
The group members could:
- Share updates on travel; restaurants to visit; events happening across Hampi, etc
- Meet interesting people
- Try to get bulk discounts from some vendors, especiallyAuto Drivers. If you have been to Hampi, you know the problem
- Help each other out with queries.
So far so good? Well, wait for it.
How (not) to run an MVP
My Idea for MVP was straightforward. I created a page on this website; Summarised the basic idea about the group and the link to join the group on the page; and ran google ads for people searching for Hampi travel/hotel booking/ and related keywords. After running ads in Bangalore and a few other cities for a week, I had absolutely zero conversions. No one joined the group.
So, I changed two things. I removed the geo-restriction of the ads and changed from the Telegram group to a Whatsapp one as it is more universal. After this change I did see significant conversions; 4-5 people were joining the group each day. But absolutely everyone who joined was some kind of spam account.
Some people joined, forwarded a few affiliate links to the group, and exited. Others were promoting their own scam websites. And some others found joy in sending absolutely vulgar voice notes to the group. The whole thing was one big mess.
At this stage, I kinda didn’t do much. Maybe call it inertia, distractions from other things, or procrastination; I just let this situation brew for quite some time without changing much. In retrospect, I should have at least shut down the ads, would have saved some money at least.
Finally, just a few days before the actual trip, I called up a friend who works at Google about the spam problem. He pointed out a silly mistake in my ads campaign which was causing the problem. But now, the time (and my ads budget) was gone. I was too close to the actual trip and repairing the campaign settings wasn’t going to get me any results.
Alas, the experiment was a failure. And not because the results were bad; bad results also mean the experiment is successful as you gain knowledge from it. The experiment was a failure because it was inconclusive. And it was inconclusive because I did not execute it well; I was not proactive and disciplined enough to solve the spam problem in time and lost the opportunity because of that.
- Respond to what is not working asap.
- Get advice from the expert early on rather than only when things break. It can save a lot of potential pain
- Document what you want to achieve, and the assumptions you want to test.
Will I revisit the solo traveling problem again? most probably. Will I do it better the next time? I hope so.