This article originally appeared on Quora: What emerging technologies are likely to be mainstream within the next ten years?
It’s no surprise that I believe AI will have a profound impact to the world we live in in the next decade. As an early stage B2B investor, I’ll examine how AI will impact the areas that I invest in heavily.
One analogy I like to use is to compare enterprise software to the evolution of autonomous vehicles. The AV world uses a level system that defines its state of autonomy (level 5 being completely autonomous). I believe enterprise software will follow a similar path.
Today, enterprise software is largely at the “power steering” phase, where workflow-based software helps you “steer” more easily. If you’ve ever driven an old car, it’s actually quite hard to steer without power steering, so the technology is valuable.
Over the next decade, I believe enterprise software will get to level 4/5, where software will be self driving, and we’ll see a paradigm shift in the coming years when we move from a mindset of machines are assisting humans to humans are assisting machines.
What’s an example of level 4/5 autonomy for software? Let’s take Salesforce for example. Salesforce has been a largely workflow driven solution to push sales reps to input their activities (so they get paid) and thus allow sales managers to view activities of their direct report and manage more efficiently.
What could a self-driving Salesforce look like? On the sales rep side, input of activity could happen automatically. The system may source and prioritize leads that have high likelihood of closing, automatically draft correspondence for these leads, and then reach out to them in the most appropriate channels (chat, email, etc). Then it’ll go back and forth with these leads to drive them down the funnel. A human may get involved when the machine is uncertain or when it’s time for the sales rep to take the potential customers out to dinner.
The fuel to climbing the self-driving ladder is high-quality, preferably proprietary data. What is proprietary data (to me)?
1) Data set is truly unique. I believe unique data sets are increasingly rare. Examples are population data (in healthcare) or time series data (i.e. data about a person over a long period of time).
2) Scale of data is proprietary. For example, LinkedIn has one of the largest resume books in the world. Is each profile individually unique? Not necessarily, but the scale is proprietary.
3) Weight of data network relationships is proprietary. Facebook has profiles, and each profile is interesting, but what’s more interesting is the weight of relationships between each person.
One of the biggest problems facing startups is how to build a proprietary dataset, and how to acquire user on day one. A few tips to keep in mind:
1) Provide significant incremental value for the 1st customer. On day 1, I believe a startup needs to provide significant value without a massive investment from the customer. This is incredibly important and I believe is a huge barrier for companies I see today. The compelling business use case on day 1 is your ticket to entry!
2) Network effect: N+1 customer gets more benefit than customer N (because of data contribution). As you add more customers, your data set should become more robust, and thus you should be able to deliver a better solution for ALL customers.
3) Get your customers to serve as a mechanical turks. In a perfect world, a startup can provide a solution where their customers will serve as mechanical turks to increase the quality of the data set without costing them additional $.
I do want to re-iterate that I see startups focus too heavily on the end goal and don’t account for the value proposition of their product today. We always have to remember that from day one, the customer is buying into what you immediately provide.
- What do the best entrepreneurs do that really gets your attention as a VC?
- What advice would you give to your 18-year-old self?
- Is AI over-hyped in 2017?