Topline · 26 Feb 2026 · From the week of 23 February
SPOTLIGHT: Raising $55M When You're Not "AI-Native" | Nick Turner, CEO @ Dreamdata
These are notes on the conversation, checked against its transcript. The episode itself has the full discussion.
In brief
Nick Turner, CEO of Dreamdata, joins host Sam Jacobs for a Topline Spotlight episode on raising a $55M Series B in a market that favours AI companies. He describes moving from CRO to CEO just as the raise began, talking to 73 investors over about six weeks, and the scrutiny on revenue durability at fast-growing AI companies. The second half covers Dreamdata's AI product work, where he argues that accuracy and verifiable answers matter more than AI branding because customers take the numbers to their boards.
For founders
- Investors are questioning revenue durability at fast-growing AI companies, asking whether the spend comes from an innovation budget or a permanent line item.
- Nick talked to 73 investors over about six weeks to close a $55M Series B, and said that closing one of 73 counts as doing a great job.
- He chose investors with realistic growth expectations, arguing that growth rarely accelerates as a company gets bigger.
- He advises against putting AI in front of a product name and says to sell the benefit and the accuracy of the output instead.
- Burn multiple and retention were the efficiency metrics Dreamdata prioritised while growing without adding headcount.
For revenue leaders
- Show users the SQL query behind an AI-generated answer so they can check that the result matches what they asked.
- Nick described a semantic layer that stops the agent from giving unreliable numerical answers, and agreed with Sam that templates and coaching help users who do not know how to prompt get the right answer to the right question.
- Because customers present measurement numbers to boards, Dreamdata approaches AI carefully so its outputs are accurate and credible.
- Nick cited a study he could not confirm was published, which found Agentforce agents failing 40 to 70% of the time, a rate he said is not acceptable.
What was said 15, most useful first
Investors are questioning how durable revenue is at AI companies that reached $100M or $200M in under a year. Listen
Nick said investors he spoke with worry about revenue durability at AI companies that have grown to $100M or $200M within a year without going through a full renewal cycle. They also question whether the budget comes from an innovation budget or a line item. He said investors still cannot stay away from such fast-growing companies, and that he was not sure investors would voice these concerns in public.
“companies that have gone to 100 million or 200 million in under a year they've not been through a full renewal cycle.”
Listen to the episode Fundraising & investors Link to this Report a problem
Nick talked to 73 investors over about six weeks while raising a $55M Series B. Listen
He compared the process to sales, where a 25% close rate is decent in a good quarter, and said that closing one of 73 investors is a strong result. Dreamdata raised the round with PeakSpan. The raise happened within months of Nick becoming CEO.
“we talked to 73 investors over the course of 6 weeks, something like that.”
Listen to the episode Fundraising & investors Link to this Report a problem
Growth rarely accelerates at larger scale, so he chose investors with realistic growth expectations. Listen
Nick said he did not want an investor who asked for 5x growth in 2026 when Dreamdata had just done 2.5x. He argued that a company does not usually accelerate its growth as it gets bigger, so he wanted a partner with realistic expectations for future growth.
“When you get bigger, it doesn't it's not going to accelerate. It doesn't usually”
Listen to the episode Fundraising & investors Link to this Report a problem
Putting AI in front of a product name is weak marketing, because buyers need the benefit rather than the technology. Listen
Nick said almost every company is putting AI in front of its product name, which he called a poor idea for marketing. He said Dreamdata has done this too. He argued that a product should sell the benefit and provide accurate data that customers can use to decide where to spend marketing dollars.
“You're not selling the benefit, you know, you're just saying AI accounting or AI, you know, marketing attribution.”
Listen to the episode Positioning & marketing Link to this Report a problem
Generative AI layered on complex data models can cause serious errors when users trust the bot's output without checking it. Listen
Nick said he has seen teams damaged by using generative AI on top of complex models after they made mistakes by trusting what the bot said. Because Dreamdata's customers take its numbers to their boards and C-suites, he said the company approaches AI carefully so its output is accurate enough for decisions.
“I've seen teams wiped out utilizing generative AI on top of these complex models completely because they made mistakes by just trusting what the the the the bot saying.”
Showing users the SQL query behind an AI answer lets them check whether it matches what they asked. Listen
Nick described an approach where the AI turns a natural language question into a structured database query, and that query is then shown to the user in a readable format. He said this involves a semantic layer, though he noted that product is his weak spot. He said the answer should not be a black box of text question and text answer.
“you take the SQL statement that you're using to query the database and then you show it to them in a format that they can understand.”
Burn multiple and retention were the metrics Dreamdata focused on to keep growth profitable and efficient. Listen
Nick said profitable, efficient growth is core for Dreamdata. He said that since he joined at about 50 people the company has grown roughly 3x while staying at 50 people, and that bringing down its burn multiple and focusing on retention were the priorities that shaped the fundraise.
“getting our burn multiple down, things like that were really important to us. So we focused on those metrics, we focused on retention.”
Listen to the episode Metrics & finance Link to this Report a problem
Many investors shy away from MarTech, so picking one that focuses on the category mattered. Listen
Nick said PeakSpan's focus on MarTech made it a good fit for Dreamdata. He said there are around 18,000 MarTech companies, so there are many opportunities, but founders need to sort themselves out within that crowd to do well in the category.
“A lot of investors will shy away from it.”
Listen to the episode Fundraising & investors Link to this Report a problem
Accuracy risk in AI analytics comes from hallucination and from users asking the wrong question. Listen
Nick said Dreamdata focuses on two risks. The first is hallucination, where he said there has been a lot of progress with large language models. The second is users who cannot prompt well enough, which can produce a right answer to a different question than the one asked.
“how do you make sure that you don't give the the right answer to the wrong question?”
GenAI products need more upfront education, since many users do not know how to prompt them. Listen
Nick said products need more education up front so people know how to use them. When Sam summarised Dreamdata's approach as layering in GenAI for natural language while keeping answers accurate, and using templates and coaching so users get the right answer to the right question, Nick agreed that this was correct.
“there needs to be more educational up front still and how people know how to use the use that as a as a product.”
A study Nick recalled found Agentforce agents failing 40 to 70% of the time, a rate he said is not acceptable. Listen
Nick said he was not sure whether the study was published, but recalled that a study done over the summer found the agents failing 40 to 70% of the time. He said that is not a place Dreamdata can be, and that answers from AI products need to work well because they answer with confidence.
“the agents were failing like 40 to 70% of the time. And that is not a place where we can be.”
Dreamdata is focusing on the application layer and waiting for LLM providers to simplify agent building, rather than building infrastructure itself. Listen
Nick said everyone is building their own agent, but Google, OpenAI and other major LLM providers are making it easier to build one. He said a company could spend a lot of resources on that work, but Dreamdata would rather be patient, noting that the major providers spend billions that a $55M raise cannot match.
“you can be patient and wait for them to make it a lot easier for you.”
Large enterprises will likely keep using external specialists for agents, since in-house agents across thousands of staff could give inconsistent answers. Listen
Nick said he thinks it could be a nightmare at a company with 10,000 to 15,000 employees if each one builds an agent, because colleagues might get different answers. He said low-code and no-code has been discussed for years without taking off in large enterprises, and companies still go to outside specialists for core competencies. He said some in-house building may happen.
“they still go to external people that have a core competency to to build this stuff.”
As a first-time CEO, he relied on strong leaders in finance and product to cover his own weak spots. Listen
Nick said his own weaknesses are finance and tech, particularly product. He said he has strong leaders in those seats who are also willing to help him learn, and that he needed someone who could both lead the function and teach him while he worked out how to help them in their roles.
“I'd say finance and tech absolutely.”
Listen to the episode Leadership & culture Link to this Report a problem
The E-Myth frames the question of whether to work on a business or in it, which shaped how Nick approached process. Listen
Nick said the book is aimed at professional service owners who must decide whether to work on or in their business, and it describes systematising a business so the owner can step back. He said he used its principles earlier when he owned a construction company and later to build process at a tech company.
“you have to decide do you want to work on your business or do you want to work in your business?”
Listen to the episode Leadership & culture Link to this Report a problem