I spent 2013 as one of four partners at a seed stage fund with a strong data focus*. During that time, I had the privilege of talking in depth with hundreds of founders about the data-centric companies they were working on. Here are some of the high level trends that I observed among those startups, along with a few suggestions for anyone who is looking for startup ideas:
Analytics is still a very hot area. Startups like Heap Analytics are rethinking the storage side while companies like Periscope are rethinking the querying side. Retail analytics (i.e. data tracking for physical stores) is getting a lot of attention from companies like Bay Sensors, Nomi, and Euclid. Panorama Education is helping schools improve through better survey data analysis. Parse.ly offers an analytics suite for publishers. The list goes on and on.
On the positive side, if you're searching for business ideas, this is a ripe space to look. On the negative side, there's a lot of competition both horizontally and in specific verticals. It's hard to get funded if there are four other companies that look just like yours, so make sure you either have a true advantage over your competition, or you pick a vertical that is still underserved.
There's a lot of value in bringing neglected industries into the 21st century. Some examples: AuditFile for streamlining audits; Sourcery for restaurant food sourcing; SupplyHog for purchasing building materials; Asseta for buying and selling industrial equipment; LendUp for making payday loans better; SimpleLegal for managing corporate legal spend; Standard Treasury for improving access to banking infrastructure; FarmLogs for managing farm operations; and so on.
These ideas don't sound glamorous, but they have the potential to be huge, from both a financial perspective and a social good perspective. Just imagine how much time and money would be saved if people could initiate wire transfers from their smartphones instead of having to work with tellers at local bank branches. Or if large companies, which currently hire hordes of auditors to analyze their legal bills and flag suspicious charges, could instead use software to manage their legal budgets more efficiently and more accurately.
In my opinion, this is the area that's easiest to target if you're looking for a good business idea. Look for situations where you're stuck doing something using pen and paper instead of mouse and keyboard, or where you're forced to use software that was made 20+ years ago. If you don't experience such frustrations in your personal life, go ask people that you spend time with: your dentist, your accountant, your barber, your librarian, your best friend — anyone you can think of.
Lots of in-house tech infrastructure is to a SaaS model. You can now host your API status page using StatusPage.io, link your customers' usage data with your CRM using Whalr, track your KPIs using Metric Insights, and create dashboards for your customers using Keen.io. When you ask most of these startups who they're competing against, they'll tell you that their customers used to solve these problems with in-house software, but now that software can finally be outsourced so they (the customers) can focus on their own core products.
If you're a software engineer, try to recall if there is any software that you had to write from scratch at more than one company. For example, I wrote rule-based data transformation tools at both of my last two jobs, which suggests a possible SaaS service (if it doesn't already exist). Alternatively, instead of thinking about software you repeatedly built, think about software that you wanted to build in-house but couldn't because you didn't have enough spare time.
Data Science tools are becoming more accessible. Data scientists write R code and use Tableau while business analysts live in Excel. The problem is that there's a wide gap between Excel and Tableau. Fortunately, startups like DataPad, Wise.io, DataHero, and Trifacta have been building tools that fill this gap. Good data analysis is extremely valuable, and it's exciting to see software that make it more accessible for non-data scientists.
If you've worked on data-oriented tasks where existing tools are inappropriate for anyone who doesn't have a statistics PhD, ask yourself if you could make a better tool that's usable by a larger pool of people. If you can, there's probably a market for it.
Those were the four biggest trends that I observed, but there are also a few smaller ones that will likely become more significant in 2014:
- A focus on privacy and anonymity with various Bitcoin startups, Whisper, and Snapchat.
- Multiple platforms and specific products in the Internet of Things space.
- Quantified self.
So that's my recap for 2013. I'm very excited to see what 2014 will bring!
* My fund partners and I believe that proprietary data offers tremendous competitive advantages, so we focus on startups that are either building valuable datasets or are building tools that help others extract more value from their data.