Pricing a new product is one of the most challenging exercises in product development. For Vizory it came down to a handful of factors that didn't all line up: what the market maths required given a limited addressable market, what directors expected based on what they were already paying for AI tools, how to structure a model for directors buying seats across multiple boards, and what to do about a pilot price that had become a reference point before the full release arrived. SaaS pricing rarely has a clean answer. This was no exception.
Testing propensity to pay alongside feature priority.
I ran a formal pricing survey with my director network that did two things simultaneously: tested willingness to pay at different price points, and asked directors to rank which jobs mattered most to them. This combined approach was deliberate — you can't price in isolation from value, and you can't understand value without understanding what people will actually pay for it.
The feature priority results were clear and consistent. Three jobs ranked significantly above everything else:
- Key signals from the pack — what are the things I might be missing that could be a risk? What should I focus on through all this noise?
- Searchability across packs — I'm sure I heard this before, I'm sure they were meant to have delivered this by now. Why have they stopped talking about it? The ability to search across previous packs using AI to find that information.
- Meeting preparation — synthesising all the notes I've reviewed, pulling out key themes, matching them to the board agenda so I can get across my key points in the meeting itself.
These three priorities mapped almost perfectly onto the features that had generated the strongest reactions in the user sessions. That alignment — between what people said they'd pay for and what they actually responded to in testing — was one of the more satisfying validations of the research approach.
The anchor I didn't see coming.
My original pricing target was $300/month (per board seat). The value case justified it: liability context, security posture, time saved. But the price also came from a structural reality. The total addressable market for a tool like Vizory is inherently limited. When you work through the SAM and SOM — directors actively seeking tools, on the right kind of boards, with sufficient pack volume to feel the pain acutely — you're not talking about a mass market. In a niche, you need a price that works at lower volume. A $20 (USD) product needs tens of thousands of customers. A $300 product can build something meaningful with a fraction of that.
The per-seat model shaped the number further. Directors who sit on multiple boards buy a separate seat for each, which is the right design: you want each company's insights isolated. But it means leaving enough margin to offer a meaningful multi-seat discount. Directors on three boards expect a break on their third seat, and if your base price doesn't have room for that without eroding unit economics, you're in trouble. Pricing for a multi-seat model means working backwards from what the discount needs to be, not just forwards from what the value is.
So the maths pointed to a premium price. The market, it turned out, had its own ideas.
The anchoring problem came not from the price itself but from the comparisons people made. For many directors the reference point was $20 a month: ChatGPT, Claude, free Gemini. Vizory is none of those things. It's purpose-built for governance, privately hosted in Australia, with persistent memory across board cycles. The cost structure and liability context are completely different. But anchoring doesn't care about structural differences. People compare what they can compare.
I made that problem worse by running a pilot at $150 to build proof of value before moving to full price at release. The pilot worked, but $150 became the new reference point for beta users, some of whom expected it to stay there or come down further. By the time of full release, ChatGPT Pro was at $120/month — a higher anchor than the free tier, but still pulling the wrong way.
I'd actually done the textbook thing here: communicated upfront that the $150 was early-access pricing and that the rate would move to full price at release. That's what the standard advice tells you to do. It didn't matter. Once a price is in someone's head, knowing intellectually that it's temporary doesn't override the anchoring effect.
The real lesson: if your full price needs to be substantially higher than your pilot price, transparency about the increase isn't enough. The only way to genuinely avoid the anchoring trap is to either launch the pilot at the full price with a credible value case from day one, or run the pilot at a price close enough to the full price that the move up doesn't feel like a punishment. A pilot at $150 with a path to $300 was always going to feel like a 100% increase to the people in it — regardless of what they'd been told upfront.
Interestingly, the directors most willing to pay a premium were also the most articulate about why they needed something other than a public AI tool. In one demo with a fund manager on multiple portfolio company boards, the opening comment before I'd shown a single feature was: "We've all got ChatGPT Pro subscriptions meant to be secure, but there's considerable hesitancy about using those for board materials." Those directors understood the liability argument intuitively and didn't need to be sold on the price. The catch was that they were also the most likely to need institutional sign-off, which pulled the buying motion back toward enterprise — a different problem entirely.
The real test came at the end of the pilot, when directors had to decide whether to convert to paid — and when I asked how disappointed they'd be if Vizory disappeared tomorrow. By that point most had been through two or three board cycles, and the price conversation had largely faded: they'd seen the value. Both questions cut through anchoring and comparison shopping in a way a pricing survey never can. One forces a judgement about loss rather than cost. The other forces a real commitment rather than a stated preference. The directors who'd been using Vizory signed on, and said they'd be very disappointed to lose it. That was a pretty important signal.
In a niche market the maths require a premium. In a trust-based market the anchor you set matters. And the real test of whether your price is right isn't the initial sale — it's whether people fight to keep using it.
Building the financial model.
I worked with an experienced finance professional to build a proper financial model, one that captured unit economics, cohort behaviour, churn assumptions, and the revenue milestones needed to reach viability.
This exercise was clarifying in ways that gut-feel projections aren't. When you build a model properly with real assumptions you have to defend, you can't hide from the numbers.
What the model showed:
- At $150/month, I needed a specific subscriber count by specific time points to reach break-even — and those numbers were achievable, but required a reliable acquisition engine, not just word-of-mouth
- Churn was the variable that mattered most — more than acquisition rate, because director-market churn is slow to manifest but hard to reverse once it starts
- Annual pricing had better unit economics on paper — but was essentially unsellable in practice. To commit annually, directors needed to trust the product. To build trust, they needed multiple board cycles. And multiple board cycles, as I've described, took months. Nobody was going to pay for a year upfront before they'd seen the product work across two or three of their actual meetings. Annual pricing is a good retention tool for established or cheaper products, not an acquisition tool for new ones in slow-cadence markets.
- The model also revealed how sensitive the business was to the length of the sales cycle — every week of delay between interest and activation had a compounding effect on the path to break-even