What’s the best way to size the TAM: top-down vs bottom-up for investors?

[Kenneth.L]

I’m trying to size the total addressable market for a niche B2B software product, but I’m stuck on whether to use a top-down industry report figure or build a bottom-up model from my own lead generation data. The top-down number feels too broad and not actionable for our specific positioning, while my bottom-up estimate seems frustratingly small. How have you all reconciled this gap when presenting to potential investors?

[William48]

In our fundraise I ended up showing a TAM range by pairing a top down anchor with a bottom up build. The top down gave a sense of macro upside, but I called out the assumptions and kept the band wide so it didn’t feel like a single number. The bottom up started out looking tiny, so we added adjacent niches, potential add ons, and an expansion path to fill out the line. Investors seemed to accept the tension when we also showed a three year plan and some guardrails for ICP reach and win rate.

[NicholasHL]

We tried to lean on our lead gen data, but the pipeline is lumpy and a few big deals polluted the signal. So we built ARR paths per ICP and then priced in expansion scenarios, cross sell, and the share of customers we expect to scale. It still looked small, but when we presented a couple of scalable tiers and a credible cadence for expansion, it wasn’t absurd anymore. What would you do in the same situation?

[Steven31]

A mentor challenged us to separate signal from noise. They asked if the real bottleneck is demand or something else like product usability, sales motion, or retention. I left wondering whether the TAM talk was masking a deeper issue and whether the pitch should pivot toward a clear value path instead of a size argument. Sometimes the tension itself starts the real conversation, not the numbers.

[Edward_T]

I laughed at myself after the last pitch when I realized I drifted into a side topic about deck layout and color schemes before circling back to TAM. The point is, I added a quick sensitivity table with three paths and a note that the base case assumes X, but it felt more like lore than a plan. Still, it gave us permission to discuss risks openly without pretending one number would do all the heavy lifting.