Reflecting On Our First 6 Months: Customer Conversations -> Change
The first six months of Hindsight have been full of ups and downs, highs and lows, laughs and conflicts. But mostly, this period has been full of change. From changing our office location, to adding an engineer, to adding/replacing innovation partners, and finally to pivoting our product, change has been the only constant in our company.
But the change has been positive, driven from appropriate decision making and purposeful market discovery. I wanted to take some time to reflect and document the process of our most impactful change, pivoting to our current product offering, Smart Tagging.
Hindsight was founded to improve the way readers discover information on topics of interest to them. We initially approached this problem from a historical perspective. How can we allow readers to better trace a story through its past key events? We pitched this tool to publications through one-on-one meetings, attending Meetup events, and participating in publishing industry conferences such as ONA and Journalism Interactive. The conversations we had with publications led to a realization and our first pivot.
Most publications we spoke with at conferences last fall were interested in the use case of our Timeline Tool, but also were interested in the underlying technology behind the tool: automated keyword/entity extraction and automated related content generation. This was shocking to us because we thought most publications adopted automated tagging and related content years ago when the technology was developed.
We discovered, though, that most publications outside of the top ~10 digital outlets still manually tag keywords and related content. As a result, we made a strategic decision to sell tagging and related content tools separately, because we would still need to develop those tools to power our Timeline product.
So we compartmentalized our products and decided to rebrand ourselves as a data science provider for the media industry. Using this pitch, we continued conversations with publications. By the end of 2017, we landed our first two innovation partners. We started building our first tool, our entity extraction and keyword tagging algorithm. But some more in-depth market research and further conversations with publications had us scratching our heads again.
We quickly realized keyword and entity extraction is a very low dollar product. Considering competitive solutions, including those from Google and IBM, the amount of money to be made on these tools was minimal, unless used in very high volume environments. Most publications did not meet this volume threshold. Therefore, we quickly determined that keyword and entity extraction was a good way for us to get our foot in the door with media outlets, but it was not a strong revenue source.
Moreover, in conversations with more publications, I found myself pitching multiple products when explaining the Hindsight offering set. By positioning ourselves as a broad data science provider, I often found myself explaining the four different products we were developing and therefore telling a story that was fragmented and too busy.
It got to a point where one customer said, “It seems like you guys are doing a lot. What is your true focus today and how can you execute on all these product with such a small team?” We found ourselves trying to do many things well without the resources to able to execute on such an ambitious undertaking, and the market caught on to this flaw. We had to take a step back and really ask ourselves what is our mission and what do we want to build. It no longer made sense to build multiple products at once. We needed to build one product first and build it well.
With the resounding feedback of sounding distracted and tackling too many products at once, we had to scale back and narrow our focus. And so we looked internally and went back to our roots to focus on the question we started with: How do we bring better context around articles to readers to improve their understanding of current events?
Consequentially, we re-harnessed our energy into creating the Timeline Tool. But before diving into the product, we wanted to further assess market fit and spoke with a few more publications. The common feedback was that the Timeline Tool is a useful resource but only for 50% of a publication’s archive, given that not every story needs a timeline. The Timeline is not useful for Lifestyle articles, “Top Ten” articles, Travel articles, etc. Publications would find greater use in a tool that could be used across all of their content. So we needed to find a common denominator. And we did.
The common denominator we discovered were entities. Every article has entities that can be better contextualized. A political story includes politicians, institutions, and locations, a lifestyle article has brands and celebrities, a business article has companies, executives, and industry terms. Regardless of topic, every story has entities that can be connected to further information. We are building Smart Tagging around this concept.
Through our Smart Tagging system, we enable readers to gain further context on specific key terms in an article with a focus on recommending news (about those key terms) from your publication. Our tools engage with readers at their peak interest levels, providing a seamless experience to better understand the news and dive deeper on topics of interest. In summary, we are enhancing hypertext to be automatically embedded, and to provide richer coverage around specific aspects of an article.
Strategically, Smart Tagging incorporates our Entity Extraction tool, which we spent the first two months of 2018 to build out. So our existing software is not going to waste. See the demo here. We will be working with our partners to release the Smart Tagging product in the next few months, so stay tuned to see how we make the internet more connected and make text richer and more dynamic!
Our journey so far has been full of hills, curves, sharp turns, and speed bumps. And I am positive the twists and turns have just begun. But it’s these obstacles that have allowed Hindsight to grow into the business it is today. We have used our learnings to develop a product that has a much stronger value proposition and market fit than our prior product visions. Here’s to continuing to maneuver through new issues and writing another long-winded blog post in another six months.
Originally published at medium.com on March 28, 2018.
