# Building a Forecast That Actually Drives Decisions

> A NachoTuesday interview — the C-Suite Insights series from NachoNacho, the B2B SaaS and AI
> marketplace. In conversation with Tom Matta, Director of Partnerships, Forecastr.

- **Page:** https://nachonacho.com/nachotuesday/video/Building-a-Forecast-That-Actually-Drives-Decisions
- **Guest:** Tom Matta, Director of Partnerships, Forecastr
- **Guest on LinkedIn:** https://www.linkedin.com/in/tommatta/
- **Recorded:** 2026-06-09
- **Watch:** https://www.youtube.com/watch?v=Bl45YIsqv-A

## Summary

This interview features Tom Matta from Forecastr and Andy from NachoNacho discussing the critical role of forward-looking, scenario-based financial forecasting for startups. Rather than treating financial modeling as a static backward-looking reporting exercise, they highlight how modern platforms integrate real-time data, human oversight, and AI to turn forecasts into dynamic decision-making tools. Tom shares vital strategic advice on navigating market volatility, balancing optimistic projections with operational realism, and regional fundraising dynamics, emphasizing that a robust financial plan is essential for maintaining investor confidence and scaling efficiently. Key Takeaways Decision-Driven over Reporting-Driven: Traditional forecasts fail because they only report past data. Effective forecasts must be forward-looking, scenario-based models (e.g., modeling a $10\%$ drop in conversion) that allow leaders to make proactive decisions before problems arise. The Power of the 90-Day Rolling Forecast: In highly volatile or rapidly changing markets, long-term $3$-to-$5$-year models quickly become obsolete creative exercises. Startups should prioritize a $90$-day rolling forecast with monthly updates to maintain high accuracy and strategic agility. The "Three-Legged Stool" of Modern FP&A: Relying solely on AI for financial forecasting is risky due to potential hallucinations. The most reliable financial planning approach combines collaborative software, generative AI (like Forecastr's Claude integration), and human analysts in the loop to guarantee $100\%$ data accuracy. Anchor Costs to the Base Case, Not the Upside: Founders should build their hiring and operating expenses against conservative base-case assumptions rather than optimistic, "hockey stick" upside scenarios. It is significantly easier to accelerate hiring if revenue outperforms expectations than it is to execute layoffs if it misses. Align Capital, Customers, and Geography: Success often requires navigating different regional ecosystems. Startups may build in one hub (like Silicon Valley or Hawaii), raise capital in another that better understands their niche (like New York for fintech), and sell to customers located in a third.

## Full transcript

Andy: [00:11] Good afternoon, everybody, and welcome to another Nacho Tuesday. Today I have Tom here with Forecastr. Without further ado, we’d love if you could introduce yourself and tell us what you guys do over there. Tom: [00:21] Hey, Andy. Yeah, thrilled to be here. Forecastr is a financial planning platform built specifically for startups. It gives founders and operators institutional-quality financial models without needing a full-time CFO, covering everything from fundraising models and ARR forecasting to scenario planning and investor reporting. The platform is used by hundreds of early-stage companies from pre-seed through Series B. It integrates with the tools founders already use—your accounting systems like QuickBooks Online and Xero, your CRM like HubSpot and Salesforce, and your payroll—so the model stays connected to real data rather than living in a vacuum. On the personal side, I joined Forecastr a little over a year ago. Before this, I was early at Carta, which is the cap table management platform. At Carta, we were taking cap table data out of spreadsheets and putting it into software because once it is in software, it’s much more manageable in terms of analytics, benchmarking, accounting tools, and reports. Forecastr is doing a similar thing with finance data. It takes your company's finance data out of spreadsheets—which typically get big, bloated, have 50 tabs, circular references, and "if-then" errors—and imports it into modern software. This makes it more manageable for analytics, benchmarking, and AI integrations. We actually have an Anthropic integration where you can use Claude to access your data on Forecastr, which completely turbocharges your finance function. Andy: [18:04] Yeah, and you're based out of Hawaii primarily, right? Tell us about the tech scene out there and the work-life balance you guys get. I’m kind of jealous right now. Tom: [01:50] So, I'm recording this from Virginia currently, but normally I'm in Honolulu, Hawaii. My wife and I were in the Bay Area working in the tech industry for about 10 years, and we started family planning. Over COVID, Hawaii was our escape. We would go to the Big Island, Maui, and Oahu. We fell in love with this small town called Kailua on the windward shore of Oahu. Because we were both working remotely for tech companies, we decided to move there and give it a go. We moved to Hawaii four years ago, and it's a very special place. There is a small tech industry in Hawaii. There are two unicorns: one is called Privateer, which is a satellite company based in Maui that sells geospatial data, with Alex Fielding as CEO. The other is called One Brief, which uses AI tools for military command and control structures. It streamlines the documentation and planning processes for divisions and battalions, cutting a process that might normally take months down in half. Additionally, there is an incubator called Blue Startups, which is a top-20 accelerator in the whole country based in Honolulu. What’s cool is that if you get into the program, they require you to move to Hawaii to live there for a few months to take the check and participate in the mentor community. Then they have demo days in Honolulu, San Francisco, and New York, so it's still very much plugged into the mainland. It’s a really special place with a lot happening in tech. Andy: [03:28] Very interesting. I feel like they don't really have to twist people's arms to get them to go try Hawaii out for a few months to see how it works out. Tom: [03:32] Exactly. What have you got to lose? Andy: [03:37] Exactly. Why do so many forecasts fail to actually influence strategic business decisions? I feel like a lot of people just look at their numbers but don't really understand them or get to that level of knowledge. Tom: [03:47] Most forecasts are built as a reporting exercise rather than a decision tool. They answer what happened instead of what we should do next. They live in a spreadsheet that only the finance team touches, meaning leadership never develops the habit of consulting the model before making calls. Forecasts that are too complex or too far from reality lose credibility fast, and once credibility is gone, people stop using them. Andy: [04:12] That's a good point. What are some of the other common mistakes that companies make when they are building or managing their financial forecasts? Tom: [04:17] A lot of the mistakes are related to prioritizing the reporting function and looking into the past versus looking into the future. One of the things we were chatting about for AI projects using our platform is how you forecast token usage or AI credit usage. There is a famous concept called Jevons' Paradox, developed by an English mathematician in the 1800s. He realized that as the technology to utilize coal became more efficient and the unit price of coal decreased, total spend actually increased rather than decreased. This happened because the lower price point unlocked so many new efficiencies and use cases. We're seeing something similar with AI token projects. In the last four or five years, the price per million tokens has decreased by 98%, but overall token spend has increased tenfold. You see stories in the market about companies like Uber blowing through their entire 2026 AI budget in the first four months of the year, and others spending hundreds of millions of dollars on tokens because they didn't put usage caps on their engineers. Because AI is moving so quickly, the 2026 budget you built in 2025 is already obsolete. You have to update it, and doing so as an organization is much more effective when using collaborative software rather than spreadsheets, where things go offline and you have to deal with version control and "too many cooks in the kitchen." Andy: [06:01] Pretty much every builder I know is running out of cloud credits; it's become an online meme at this point. That is a great historical reference, though. What is the difference between a forecast that simply reports numbers and one that actually helps with decision-making? Can you walk us through how Forecastr helps with that? Tom: [06:21] A reporting forecast looks backward, whereas a decision-driven forecast looks forward and is built on scenarios. Decision-driving forecasts model what happens if conversion drops 10% or if a key customer churns, allowing leaders to act before problems materialize. The best forecasts have a clear owner who updates them regularly and brings them into weekly or monthly leadership conversations. When a company signs up for Forecastr, they work with one of our onboarding specialists to recreate their model and forecast in our platform. From there, your Forecastr analyst actually owns the model. They receive regular updates from the CEO and key executives—such as pipeline data from HubSpot, hiring plans, or payroll plans—to keep the model accurate. We then hold a monthly sync to review this information, ensuring the executive team has the best available data to make proactive decisions. Andy: [07:36] How should founders think about operating in uncertain or rapidly changing markets? We are one tweet away from entire markets being upended these days. What recommendations or tips do you have for founders to stay on their toes? Tom: [07:48] In uncertain markets, the goal of a forecast shifts from prediction to preparation. You are building a map of the decision points you'll hit, not claiming to know the exact future. We always suggest that our clients narrow their forecast horizon. A 90-day rolling forecast with monthly updates is far more effective than an 18-month model that is bound to be wrong by month three. It reminds me of when I was a founder 10 or 15 years ago, and investors would prep us to build three-to-five-year financial models. Trying to project what years three, four, and five look like is pretty much just an exercise in creative writing. Andy: [08:29] Yeah, a lot of those 30-page business plans are out the window now. Tom: [08:36] The game has changed. They don't do that anymore. We really recommend starting with a 90-day rolling forecast with monthly updates instead of focusing too far into the future, where the likelihood of accuracy drops to zero. You also need to stress-test your key assumptions early and often. If the model breaks when your customer acquisition cost (CAC) goes up 20%, that is a critical thing to know before it actually happens. Andy: [09:05] What kind of inputs or assumptions tend to have the biggest impact on forecast accuracy? Tom: [09:13] Conversion rates at each stage of the sales funnel can be highly impactful. A five-point swing in trial-to-paid conversion can rewrite your entire revenue line. This is also highly dependent on your business model—whether you are B2C, B2B, or an AI foundational model using consumption-based usage. For SaaS companies, Net Revenue Retention (NRR)—whether existing customers expand or contract—is often much more powerful than new logo assumptions. Time-to-revenue is also critical. How long does it actually take from contract signed to cash collected? This is consistently underestimated by founders, and it can severely throw off your cash flow forecast. Andy: [09:56] People love to get paid, but nobody likes to pay you. I used to run an agency back in the day, and I remember chasing invoices every single month. It was incredibly stressful trying to plan around that. Some clients paid promptly, but others had aggressive terms that put you 30 to 45 days out. It can also be a major problem at renewal time. Tom: [10:22] Most companies are signing 12-month contracts. The most sophisticated organizations have a process where account managers work with sales or upsell teams to evaluate where a customer is in their journey. Are they happy? Is this an upsell opportunity, or is it a churn risk? You want to identify this early so you can navigate the renewal smoothly and secure them for another 12 months, or even a multi-year contract if possible. Andy: [10:46] What role should forecasting play in aligning leadership, finance, operations, and the rest of the team behind a single financial vision and performance standard? Tom: [10:51] The forecast should be the single shared version of reality. When sales, ops, and finance are working from different numbers, decisions become political rather than analytical. Monthly forecast reviews with cross-functional attendance create accountability. If sales is projecting 40 new logos but operations hasn't planned for the onboarding load, the forecast exposes that gap before it becomes a crisis. Leadership alignment happens much faster when everyone looks at the same model and debates assumptions rather than arguing over whose numbers are correct. That is one of the greatest strengths of using an FP&A tool like Forecastr. The model lives in Forecastr as your single source of truth, and you can set custom permissions so team members see only the portion of the model that directly influences them, while executives can see the whole picture. You don't have to worry about who has the most current version of a spreadsheet; you just log in and see what you need to know. Andy: [12:02] How do you help companies move beyond static spreadsheet forecasting to dynamic, scenario-based planning? What does the onboarding process look like for Forecastr? Tom: [12:09] We always start by identifying the three or four assumptions that move the forecast the most, and we build dynamic sliders around those. You don't need to model every single thing dynamically on day one. The first step is getting the model out of email chains and into a shared tool where updates are visible and versioned. Most of the friction in spreadsheet forecasting is version control, not the model itself. From there, the organization must commit to a monthly cadence of comparing actuals versus the forecast. The discipline of comparing what you predicted to what actually happened is what makes any model get better over time. Andy: [12:42] What kind of faux pas or red flags do you commonly see when people are producing their forecasts? People probably over-index on sales they assume they are going to close, but what other red flags do you see across the board? Tom: [13:00] When actuals consistently come in below forecast on the same line item, that suggests a model assumption is wrong, not just a one-time miss. Another red flag is when cash runway is shrinking faster than modeled even though revenue looks on track; that usually points to a timing issue with collections or unplanned expenses, like spending more on AI tokens than anticipated. Additionally, if your sales pipeline coverage drops below three times your quota with less than 60 days left in the quarter, a revenue miss is already locked in before it even shows up in the forecast. This is why CRM integration is so important. For companies with longer sales cycles, you need to know if there is a bubble or a gap in your pipeline months in advance rather than reacting to it when it's too late. Andy: [13:50] That's very interesting, especially regarding renewals. If you have a large batch of renewals coming up and expect a certain percentage to drop, you could hit a massive, unexpected drop in revenue if you haven't planned to backfill it with new sales. How do you balance optimism with realism when building projections and operational plans? As a founder, you love your product and want to show a cool "hockey stick" growth chart to investors, and investors often like to see that too even if they know it’s a bit of a stretch. How do you strike that balance? Tom: [14:29] If you weren't an optimist, you wouldn't be a founder! You have to think positively, but a dose of realism is crucial for preparation. We always advise anchoring your base-case revenue projections to what is already in your pipeline or under contract. Everything beyond that is upside; it shouldn't be treated as the baseline. You should build your hiring and cost plans against your base case, not the upside. It is much easier to accelerate hiring if revenue surprises you on the upside than it is to conduct layoffs if it doesn't materialize. Optimism belongs in your upside scenarios, not baked into your baseline assumptions. That way, you aren't managing your day-to-day business to a number you secretly knew was a stretch. Andy: [15:10] That's a great point. Hope for the best, but prepare for the worst. What metrics or forecasting models are most useful for SaaS and high-growth companies today? Tom: [15:23] There are several classics. First is the ARR waterfall, which breaks down Annual Recurring Revenue into new ARR, expansion, contraction, and churn. It’s the clearest way to see the health of your revenue engine rather than just looking at a top-line number. Second is the Customer Acquisition Cost (CAC) payback period and Lifetime Value (LTV) to CAC ratio analyzed by cohort. Growth that takes 36 months to pay back isn't really growth; it's capital consumption. Finally, the Rule of 40 is still a great North Star metric for SaaS. Your growth rate plus your profit margin should ideally be above 40%. It keeps leadership from optimizing solely for growth at the expense of business efficiency. Andy: [16:07] Especially with churn rate, understanding how long people stay and where they drop off is key. How can better forecasting improve investor confidence, budgeting decisions, and resource allocation? Tom: [16:20] Investors fund the team just as much as they fund the business. A founder who can walk through their model, defend their assumptions, and explain what metrics they are tracking is far more fundable than one who just presents a static number without a story. Better forecasting also means fewer surprises at board meetings, and surprises are what erode investor trust the fastest. When your forecast drives resource allocation, every dollar of your budget connects directly to a strategic assumption. That discipline prevents the "we spent it because we had it" problem that kills efficiency as a company scales. Andy: [16:59] It's great to show that you have a plan and that your assumptions are backed by a real, strategic thesis. Looking ahead, how are AI, automation, and real-time data changing the future of forecasting and strategic planning? Tom: [17:17] AI is shifting forecasting from a monthly finance exercise to a continuous process. Models can now update in real time as actuals come in, without anyone having to manually pull data. The bottleneck is moving away from building the model to asking the right questions. Founders who know how to interrogate their data will get dramatically more value from AI-assisted planning tools. The most meaningful near-term change is scenario modeling at speed. What used to take a CFO a week to stress-test can now be done in minutes, allowing leadership to make faster, better-informed bets. Andy: [17:47] Good point. What is next for you and Forecastr? Do you guys have any exciting partnerships or features coming up? Tom: [17:53] We are doubling down on making institutional-quality financial modeling accessible to founders who don't have a CFO. The goal is for any Series A or B company to walk into a board meeting with the same financial model rigor as a public company. Our focus right now is also helping founders understand AI's impact on their P&L before it surprises them. Token spend, gross margin compression, and agentic cost dynamics are the new headcount conversation. The best next step for anyone in the audience is to get their current financial model into Forecastr and run the AI cost scenarios we talked about today. The earlier you build that visibility, the better positioned you are for your next raise. Andy: [18:31] You have spent time in several different tech hubs. What trends and insights have you noticed, especially after spending time in New York last week for New York Tech Week? By the way, Tom came straight off the plane to our event, so we appreciate him making it! How do you compare the trends you've seen across markets like San Francisco, Seattle, New York, and Hawaii? Tom: [19:00] New York Tech Week is one of my favorite tech weeks in the world. There's an incredible vibe to being in Manhattan, taking the subway from event to event, and going to venues like the NachoNacho event at Mr. Purple in Midtown. It was a killer roof deck where you could see the Empire State Building on one side and One World Trade Center on the other at dusk. The music and energy were great, and it was the perfect kickoff party for New York Tech Week. Forecastr was proud to sponsor that. In terms of regional trends, New York is an amazing tech town. I went to some events with seasoned New York investors who raised their first funds 30 years ago and are now on fund eight. It's a highly dynamic ecosystem where people are building incredible vertical SaaS, vertical AI, agentic workflows, foundational models, and direct-to-consumer fashion tech. It is a unique intersection of different industries where you can pull diverse talent to build awesome products. I lived and worked in San Francisco and Oakland during my time at Carta. I was the 12th employee there and saw it grow to over a thousand people. Carta takes equity information out of spreadsheets and puts it into software to help companies manage cap tables and let employees exercise options. Now they are positioned as the "ERP for private capital," servicing funds and private equity with $500 million in ARR. A wild story about Carta's Series A is that they actually weren't able to raise money in the Bay Area. Silicon Valley investors didn't quite grasp what the CEO, Henry, was trying to do. But they went to New York and got multiple term sheets right away because the New York market deeply understood fintech. They realized that if Carta succeeded in building a "NASDAQ for private capital," it would be massive. The lesson there is to build where your strengths are, but go where the capital and customers understand your value. Henry loved building Carta in the Bay Area, but he had to raise money in New York to make it happen. The same applies to Hawaii. I mentioned the AI military planning tool One Brief. The CEO told me the absolute best place in the world to start that company is Hawaii because so many branches of the US military and Department of Defense have command and control structures on the islands. He can sell directly to the Marines, Navy, Air Force, and Space Force right there. In terms of being close to your customers, Hawaii is a fantastic place for that business. So, be mindful of where you build, where your customers are, and where the capital is, because they might not all be in the same place. Andy: [22:48] That is a great answer. A lot of people fall into the trap of thinking everything has to be in one place—that if they are in New York or San Francisco, their customers, fundraising, and team must all be there. But like you said, you can build in one city, raise from another, and sell to a third. Don't be afraid to check out other tech hubs. Austin is another great up-and-coming one, and LA is excellent as well. We had a founders' event in LA recently, and Silicon Beach never lets me down. Tom: [23:35] When you go to an event in LA, it's such a unique vibe. People are incredibly fashionable and make bold style statements—brightly colored hair, studded leather jackets, vibrant dresses. It is highly adjacent to influencer culture. You meet founders who are active on TikTok, Instagram, and YouTube. They are getting paid to promote products, but they are also actively investing in new tech. There is a strong consumer packaged goods (CPG), advertising, and entertainment culture in Southern California that you don't get in cities like Austin or Chicago. Andy: [24:22] That's one of the reasons I love doing events in LA. You meet so many characters, everyone knows how to dress, and the vibe is very positive and happy, whereas in San Francisco or New York, people can be a bit more transactional and to the point. There are also a lot of great SaaS companies in LA that adopt that creator influence. They might run a SaaS company, but they’ll show up to your event as the most well-dressed person in the room. Tom: [24:59] Or they show up in shorts, a t-shirt, and sandals! You never know who you are talking to. I love the style decisions of people like Palmer Luckey or Palmer’s peers, whose uniform is a Hawaiian shirt, shorts, and flip-flops every single day—similar to Steve Jobs with his black turtleneck. Andy: [25:13] Usually, those are the people who have the most going on because they’ve reached a level where they don't have to care what anyone thinks about their outfit anymore. To wrap things up, what are the key takeaways people should know about Forecastr, and why should they sign up today? Tom: [25:28] I would say that if you are thinking about financial projects, there is a positive-ROI project in your future—whether it’s planning, analytics, or benchmarking against comparable companies. Schedule a call with us. Not only do we have software to help, but we also offer fractional CFOs for custom projects. Furthermore, once your data is in Forecastr, you can use our Anthropic integration to ask Claude questions about your financial data. However, remember that AI can hallucinate; it might give you answers that are correct 95% of the time, but for financial data, you need 100% accuracy. That is why we have human analysts to QA and QC the data generated by the AI. We believe the ultimate long-term solution is a three-legged stool: a combination of humans, AI, and software working together. If you are only using one or two of those three, it's not a complete solution. We would love to have a conversation, learn about your business, and help you on your journey. Andy: [26:42] That's a great point. Especially with finance, you should always keep a human in the loop to prevent errors. You don't want to run out of runway because an AI hallucinated a more optimistic scenario than reality. I heard a horror story from a founder who realized late in the game that the AI tools they were using to forecast were only 90% accurate, which led to poor decisions based on faulty information. An ounce of prevention is worth a pound of cure. If you don't have that financial expertise on your team, find a partner like Forecastr to be that human in the loop for you. We are very excited about this partnership, Tom. I'm looking forward to seeing you either here or on my next trip to Hawaii! Tom: [27:42] We need to do a NachoNacho event for Hawaii Tech Week! Andy: [27:47] Sign me up! I’ll have to convince Sanjay, but I would love to do an event out there. Thank you, everyone, for tuning in today. Be sure to check out Forecastr at forecastr.co or find them in the NachoNacho marketplace to access great software products and partner benefits. Thank you for coming on, Tom, and I look forward to seeing you soon! Tom: [28:29] Sounds good. Thanks, Andy. Thank you, everyone.

---

All NachoTuesday interviews: https://nachonacho.com/nachotuesday
About NachoNacho: https://nachonacho.com/llms.txt