Back in March, the BOQ Hackathon was held in Brisbane, and after reviewing the mentor list, I did some research about what the different mentors brought to the hackathon and if I could learn anything from them.
One of the mentors had created a website, “KanyeTexts” in a day, using a bunch of cloud services linked together; accepting payments, and sending many, many annoying Kanye West text messages to friends/enemies.
It got me thinking immediately of creating a Taylor Swift version…
Lessons Learned
- Use off the shelf products to create “minimum viable products”. Using Stripe/Zapier/Twilio, I was able to pull together a very quick solution to take payments and automate SMS messages in bulk. This wasn’t very original compared to KanyeText; the principal was the same. But for me, it was a big lesson learned. It’s the Pareto Principle in practice; 20% of work provides 80% of the value. I also discovered how useful Zapier is, and came up with many other ideas!
- Zapier costs money. And a lot of it. I ended up writing 4 Zaps, but didn’t realise until fairly late, that each activity in your Zaps counts towards your usage. This becomes very expensive very quickly; 3000 actions is roughly $50USD per month; for my platform, this equated to around 300 text messages…Twilio’s SMS messaging service is a fraction of this cost for the same number of messages. If I were to continue this service, I would be best to write my own scripts to replace Zapier; it was great to rapidly prototype, but once I proved the workflow and automation, it was evident that I could do the same thing just as easily with a bunch of scripts on my webserver. Writing that code is a to do item, but one that was validated by the prototype.
- If you’re going to copy an idea, do it better. I didn’t realise until after I spent all weekend on it, that the same author had actually written a Taylor Swift version, despite their blog post saying they wouldn’t bother; I failed to research any further; but I learned from their blog and lessons learned and applied what I do in my day to day job; automate and find efficiencies, everywhere. What resulted was a “multi-tenant” type idea where I could copy the base website, customise it with <insert celeb name> branding and text content, but use the same backend over and over.
- Have a marketing plan. Who the hell pays $100 to send 200 text messages?! Jilted ex lovers, creeps, really annoying people… I had to intervene on multiple occasions to reply texts threatening to call the cops as a result. I had no idea who would want to use this service, so instead of advertising it, I added a bunch of my contact list’s phone numbers to the database and hit send; momentarily hilarious, followed by a few “wtf Ashley”‘s, followed by my friends sharing the service to their friends who shared it when they bombarded their friends and so on. Word of mouth marketing worked great here but it isn’t going to work for everything. My “marketing” plan was more dropping my friends in the thick of it and hoping they’d
What’s next?
Not more text messaging. TayTayTexts.com lasted a whole 2 weeks before I took it down, and it will never be revived. I realised, that’s just god damned annoying. I took away a few great lessons here in creating an MVP, working with webhooks and off the shelf integration, and added a few additional WordPress tools up my belt. I’ve got a few ideas up my sleeve now for other “things” to create.
Reading
This little project spawned a whole bunch of Audio Book purchases and podcast listening, the following really got me hooked and had some great lessons!