Top Down & Bottom Up in AI - A match made in......some companies!
In my previous life (before starting FernAI), doing Top Down and Bottom Up planning was a regular occurrence and part of any strategy for the companies I worked for.
The process was useful.
Top Down– What we needed to achieve
Bottom up– How we get there
The gap between the two created our planning including investments required in order to close the gap. These investments were costed, given a timeline, and exposed to all parts of the business in order to have a transparent commitment from all parties.
Everything, and I mean, everything – linked back to P&L
- Investmentrequired
- Likely returnon that investment
- Likely timelinefor that return
- What is the cost of not investing
The separate worlds of AI &……AI
What’s been incredible to me is how disconnected the strategic world of AI is from the technology driven world of AI
Here we have one of the most revolutionary technologies that has the ability to drive a real competitive edge, and yet, how it ties into our overall strategies and objectives becomes a secondary subject of discussion
Top Down:
- in AI, we accept hypothetical outcomes as opposed to financial ones
- In AI, we accept ‘because I say so’ as proof of performance instead of clear Value tracking
- in AI, we seem to have an invisible & uncontrollable amount in our annual forecasts for ‘AI wastage’, without holding ourselves or others to account
Bottom Up:
- We’ll evaluate technical solutions, against technical problems, and that process will be exhaustive
- We’ll head into programmes of work trying to connect A to B, or move from X to Y without being sure what that does to our top-down strategy
- We’ll debate the best hyperscaler, buy-over-build, the latest tech innovations. We’ll usually do this based on gut feel
This is the reality of the AI world today. 2 agendas, both designed to enable each other, but both tackled in isolation.
Imagine it was different
I’d like to suggest an exercise to anyone reading this who is on their own AI journey, and it is designed to be done in this sequence:
- Download your companies latest financial report that is publicly available
- Study the revenue growth, costs, EBITDA – even better do this over the last 3 years and see the trend
- Reconcile these results against the latest CEO report, where hopefully, you’ll find some soundbites on where they want to see the next 12/24/36 months going
- Start brainstorming where AI can make a tangible difference, and map it back to points 1, 2 & 3. Think top-line & bottom-line effect.
- Now for the bottom up. Think about how you get there…
- Are there many dependancies (maybe also described as blockers) in place? List them.
- What do you need to solve in point 6 to enable you to move forward on point 4
- Does the investment required vs timelines to execute give you in year returns, if not, is that feasible?
- Does the complexity (and thus investment) to execute, warrant the financial outcome possible?
- If you’re still in the ‘Yes’ bracket at this point, start thinking about how you set up Value tracking to prove this during Pilot & Deployment.
Now, you’re starting to bring the 2 worlds together.You’re innovating but against the backdrop of Strategy.
At FernAI, this is our approach. We come from the bottom up world, but we approach AI from the top down.
A match made in….some places. But with FernAI, a happy path!
#StartWithValue
