May 18, 2023
At our last company gathering, Angus told me (well, shouted to me over the clamour of the crowd and music, if we're being honest) about the philosophy of Always Be Forecasting. The idea immediately resonated with me, and I had to know more. A noisy restaurant wasn't the place to dig deeper, so I left it for the moment. But the idea had its hooks in me, and wouldn't go unsatisfied for long...
You see, I'm a PC gamer at heart. When playing a ranged or healer class in an MMORPG like World of Warcraft or Final Fantasy XIV, one of the first things you're taught when trying to improve is ABC: Always Be Casting. A boss fight is of finite length, so you should never be sitting around wasting your most valuable resource (time) doing nothing when you could be doing something, even if that something isn't perfect.
I wanted to understand if Always Be Forecasting was similarly about optimising resources. Our platform gives gaming companies the ability to run forecasts whenever they want and get results in minutes, so was the idea that if they weren’t doing that, they were wasting a valuable resource, and losing time they could never get back?
Angus pointed me to Jai, our Head of Client Strategy and Insight, to learn more. I caught Jai on Slack and asked him why he thought it was important that gaming companies Always Be Forecasting… and it seems appropriate that I recount what happened next in meme format:
Despite briefly considering packing in our jobs and forming a hip-hop duo to rival the likes of DJ Jazzy Jeff & the Fresh Prince (and yes, that reference dates both of us), Jai did eventually explain.
You see, it’s all about the change-confidence feedback loop. The more confidence you have around performance, the better prepared you are to make changes. And the faster you can see the effect of your changes, the more confidence you have in their performance.
Gaming companies have to be able to run A/B tests (such as testing out a strategy or feature in a smaller segment) and get quick forecast results to decide whether the change will result in an uplift in a larger market. Spending more time waiting to see the results of a test comes with a very real opportunity cost. You could be missing out on getting that change to a larger market sooner and reaping the reward, or you could be moving on to the next test instead of waiting for the results from a dud. Further, getting results that only represent the segment’s immediate response without seeing how that tracks in the forecast can lead to a lack of confidence in the result and a delay in decision-making.
This is true for soft launches too. The faster you know if and when a new game will pay back on its investment, and the more confident you are in that forecast, the faster you can decide whether the game should be improved, the UA strategy adjusted, or to pivot entirely.
The final example he gave was around budget performance. If you’re only finding out the over/under on your Q1 budget on March 31st, you’re doing it wrong. Regular forecasting will warn you much earlier if you’re not tracking according to plan, while you still have time left to actually do something about it.
The last piece of wisdom Jai delivered was that ‘Always’ doesn’t mean ‘Daily’. Once a week is good enough to give tests time to bake, changes time to reflect, and decisions time to be executed. Before a tool like ours put the ‘fast’ into ‘forecast’, even weekly forecasts would have been unthinkable! But now they’re quick, easy and automated. So if you’re forecasting less often than that, then you’re not effectively tracking how your real-time performance affects your future, and you’re not optimising your change-confidence feedback loop.
What I’ve learned from the journey is this: the gaming market moves fast, and if you want to keep up, you better ABF: Always Be Forecasting.
Also, my colleagues at Ramp are pretty excited to share their vision, and definitely passionate about it! Even if the format is a shout-talk over a beer or a rap battle on Slack, it’s easy to learn a thing or two from the people here.