Two months later, I want to share that Sapiengraph makes for a poor business

I started the Sapiengraph effort by hiring a small sales unit. Historically, I always built the product first then attempt to sell it. The risks are obvious. I might invest months on a product that nobody wants. With sales first, I let the market decide instead. Besides, knowing myself and our team, the technical part of the business is never the problem.

This post intends to share the progress in our effort in market discovery

In the last two months, the sales and marketing team, and I have been hard at work gathering feedback on three fronts:

• Sales pitches
• Fund-raise pitches
• Content marketing

With the feedback, the questions that glare back at me are:

• What exactly are we selling?
• Who are our customers?
• How large is the potential market?

I thought I knew the answers. But our outreach strategy was random at best. We were targeting AI (data analytics) companies, and we were hoping to be AWS for their data needs. But what were AI companies?

We considered Grab an AI company because they have some predictive algorithms that required data to be trained.

We also considered quantitative trading companies to be a part of our target market because they utilized algorithms and big data.

These companies needed data, and we were able to crawl data. Once we got into the nitty-gritty details, questions surfaced.

• But what data exactly were we offering that was so unique and immediately useful to AI companies?
• Why do they need us?

The dilemma - infrastructure or solution

I love building infrastructure products like Kloudsec. Infrastructure products are products like Diffbot or Amazon AWS. They allowed other companies to build bigger and cooler things with their product offering.