Interviewing Rael Cline, Co-Founder and CEO of MediaGamma — Key Takeaways
Last week we were very fortunate to interview Rael Cline, co-founder and CEO of MediaGamma. He kindly shared the story of his entrepreneurial journey, how MediaGamma was launched and his plans ahead for the company, as it is looking to scale. MediaGamma is also currently hiring.
Below are the top 3 unique insights from our interview with Rael:
Do you have any advice for B2B startups on how they should secure their first client? [17:01]
You need to be pragmatic about these things. If you are proposing an AI solution, a lot of problems don’t need to be immediately solved by AI. A lot of gains from an AI solution can be on the automation side. Do something low risk, probably back office. Demonstrate value that way. Focus on the type of task very, very narrowly. And be able to pull off those successes. You need to build trust. If a large corporate hasn’t engaged with you before, it is very hard to get datasets, often very sensitive data for you run all kind of experiments with. Therefore, start very narrow, low risk, prove your value, build up your trust.
How do you assess the opportunity cost of each [AI] project? [23:15]
Firstly it is the question of “Is this a problem that AI is uniquely positioned to solve”? Does the customer view this as something strategic? You need quite a bit of patience for these projects and the ROI of these projects. Technical considerations around data integration, data sets: is the data in good enough shape? Is it a really big point for the customer? If it works, what does the relationship look afterwards? Are they viewing it as a consultancy project or as a SaaS type relationship? Is this a problem that is shared across other types of companies?
Will you ever be done with building the ultimate AI algorithm? What would be next? [26:39]
Definitely, the short answer is no. I think it’s the wrong thing to be optimising on anyway. If you are a commercially focused organisation, hypothetically if you are 90% accurate, becoming 91% accurate is probably wrong goal to be optimising on.You really need to understand where you are going to move the needle. The second part, I think it is incredibly risky to be building your business just on an algorithm, Google has done it and that’s probably a huge anomaly - I think it is very difficult to do that. You will need to find other moats to defend your proposition.
If you are short on time, you can listen to the interview on SoundCloud.
Alternatively, you can watch the full interview with Rael below.
Here are more useful highlights from our interview with Rael:
For how long did you collaborate together before deciding to launch the company? [4:54]
Timing is such an important thing when starting a company and it is so hard to control it. Most companies are either too early or too late and it is really hard to control. But on the other hand, you don’t want to make a snap decision. You really want to get to know someone, do they have the same ambitions for the company? What does the success look like? Obviously, is there trust and have you known each other before? There was a lot of validational research to be done anyway, both of us were doing that before any formal commitment. There were probably a good few months between the first contact and a formal agreement. It was a chance for both of us to demonstrate that we were serious about the creation of an entity to commercialise it.
What kind of customer development did you initially do? How did you validate that this was a pain point for your customers? [7:02]
It’s really about narrowing down the problems, we were really quite lucky in the early days. A lot of lead generation came from word of mouth referrals. There were quite a lot of inbound enquiries for us.This clearly demonstrated demand without us having to actively market it, this doesn’t exactly translate into a big enough pain point, into something that is repeatable and scalable. It was quite a careful customer development process. With all the inbound leads, we considered if they were big enough problems. It was really about figuring out the most common problems that fit into the repeatable and scalable category.
Why should someone join MediaGamma?[25:06]
Ultimately it comes down to “Are you solving a really interesting problem?”. Are you attracting like-minded really world class people? Is there a culture which is going to be around and help build something? I think for us the quality of the people on the research side and how it then translates into the commercial setting is really important. Some of the data science work we are doing around reinforcement learning are really interesting and hard problems to solve. The group of people is already really fantastic, we have individuals from all the corners of the world. We have a common understanding of what we are trying to do. We make sure everyone feels valued. Everyone in the company can see their impact on how it is moving the needle for the business. We are able to demonstrate this and that’s why it is an environment that people want to come and work in.
