Confidence. Before the launch.

In brief: Decision intelligence is the practice of designing and testing a decision before acting on it. In marketing, that means comparing messaging, creative, positioning, pricing, or audience options before committing budget. At Galaxies, these tests run in Nexus using synthetic personas built from a client’s own real research. They do not replace market research; they bridge the gap between the data a company already has and the approval of its next campaign.

At a recent talk for marketing leaders called “The Consumer Your Data Lost,” I asked the room how many weeks had passed between the brief for their last major campaign and the day it went live. Then I asked them to multiply that number by seven.

The answer is the number of days consumers had to change their minds about the category while the company approved copy, creative, channels, and budget. An eight-week approval process gives them 56 days. That is long enough for a trend to emerge, peak, and fade.

That delay has a name: the Signal-to-Decision Gap, the time between a consumer signal appearing and a company making a material decision based on it. As long as that gap persists, brands make decisions a step behind their audiences.

I bring this up because the conversation about AI in marketing has become stuck on production: more copy, more ad variations, more images. That solves a problem that is no longer the main bottleneck. The harder question is which of the many available options deserves the company’s money. Decision intelligence addresses that question. Gartner takes a broad view, close to BI. Galaxies has focused since 2022 on a narrower challenge: making consumer decisions before spending, for consumer-facing businesses.

What is decision intelligence?

Decision intelligence is the discipline of structuring a decision before making it: define the problem, state a hypothesis, develop alternatives, test how each performs, and record the outcome so the next decision improves. The focus shifts from the data itself to the choice the data must support.

Gartner describes it as a practical discipline that advances decision-making by explicitly understanding and engineering how decisions are made and how outcomes are evaluated, managed, and improved through feedback. In its 2025 AI Hype Cycle, Gartner estimated adoption at 5% to 20% of organizations, with two to five years until maturity [1].

For marketing leaders, the important phrase is “engineering how decisions are made.” Most companies I know design campaigns with great care. Very few design the choice between campaigns. Alternatives are often compared only in an approval meeting, where the most confidently defended option wins, not necessarily the best-tested one.

How does decision intelligence differ from BI, analytics, and generative AI?

These approaches coexist in the same operation. The difference is the question each answers. Business intelligence asks what happened after execution. Historical predictive analytics asks what is likely if past patterns repeat. Generative AI asks what options can be produced. Market research asks what an audience thinks, with sample-based evidence. Decision intelligence asks which alternative to choose, and why, before execution. Its test is only as good as its underlying data and hypotheses.

Generative AI is the most revealing comparison. It multiplied the number of options without improving our ability to choose among them. A team that once debated three campaign concepts now debates thirty, with the same deadline and the same number of people in the room. Without a decision method, more options create more opinions, not better choices.

Why has marketing run out of time to make good decisions?

The first two signals show that consumers now make decisions with help from AI, often in conversations the brand cannot see. The others show how Brazilian marketing leaders view their ability to measure and defend their choices. Clutch reported that 65% of consumers use AI to research products before buying [2]. PYMNTS reported that 53% of Brazilians used AI in their most recent online purchase [3].

In a 2026 Makers and Uncover survey of 113 marketing leaders, 75% said data was a determining factor in their decisions, yet only 9% fully trusted what they measured. Another 41% said their boards judged marketing primarily by revenue growth or market share [4]. The research ran from May to June 2026 and included CMOs, VPs, directors, and managers; 70% worked at companies with revenue above R$500 million. Business results are demanded, but the evidence for defending each decision before spending is still weak.

This is not a skills shortage. These teams have creativity, brand experience, and market knowledge. What many consumer businesses lack is a way to compare options before committing budget and to arrive at an approval committee with defensible evidence.

How does Galaxies’ Testable Decision Cycle work?

At Galaxies, we organize client work into five steps, which we call the Testable Decision Cycle. The method is useful without any particular tool, although Nexus was built to run the testing step in minutes and deliver a structured, human-reviewed verdict within 48 hours.

Figure 1. Galaxies’ Testable Decision Cycle: problem, hypothesis, alternatives, test, and decision. The team brings real-world results into the next decision.

1. Problem: what decision must be made, and by when?

My first question to a client is not “what do you want to research?” It is “what do you need to decide, who signs off, and on what date?” A decision without an owner or deadline becomes a study; a study without a decision becomes a report in a drawer. We also estimate the cost of getting it wrong, in budget, time, or reputation.

2. Hypothesis: what does the team believe would have to be true?

Every campaign proposal contains an implicit bet: “Younger audiences will respond better to convenience,” or “Discounts are the strongest Black Friday trigger.” Writing that belief as a single falsifiable sentence is the step that most changes the quality of the internal conversation.

3. Alternatives: at least two options, plus what already exists

The most common mistake I see is bringing just one option to a test. Testing a single asset validates an opinion; it does not support a choice. We ask for at least two genuine alternatives and, where available, the current version as a baseline. Without a comparison, any result can look good.

4. Test: how does each option perform with each audience?

This is where synthetic personas come in. We show each option to profiles built from the client’s real data and examine three things: which option each audience prefers, why, and what would make them reject it. Barriers are often more useful than preferences because they point to what needs to change before an asset goes live.

5. Decision: choose, record the rationale, and define the next step

The test ends with a documented choice and the criteria behind it. We also decide what still needs formal field validation. After a campaign runs, the team compares the real outcome with the test’s indication and carries that difference into its next decision. Recording that learning turns a series of tests into cumulative knowledge.

Which marketing decisions can be tested before execution?

Not every decision deserves the same effort. Pre-testing is most valuable when the budget is high, the deadline is tight, or the choice is difficult to reverse. Six decisions come up repeatedly in our work: messaging, creative, positioning, launches, pricing, and audience selection.

For messaging, compare two or three ways of expressing the main promise by audience and identify the objections each raises. For creative, rank competing images, videos, or copy before buying media. For positioning, compare alternative brand territories with the existing one to understand fit and the risk of alienating current customers.

For launches, test product concepts, names, and benefits to identify adoption barriers by profile. For pricing, compare how a price, package, or commercial terms are presented and what value each conveys; this does not replace a price-elasticity study. For audience selection, expose different segments to the same strategy and identify where it resonates or needs adjustment.

Each of these decisions is part of this month’s broader series on predictive marketing, predictive analytics, consumer behavior, value propositions, brand positioning, market segmentation, and scenario planning for the 2027 budget. Links to those articles can be added as they are published.

What do the cases show about deciding before spending?

At Mahta Bio, creative validation before the campaign reduced customer acquisition cost by 35%. A decision that used to take six to twelve weeks can now receive a structured verdict in 48 hours.

That result is often treated as a media-efficiency win. I see something else: once testing fits the timeline, it stops being an annual event and becomes part of routine approvals. That change in frequency makes decision intelligence possible in marketing. The simulation did not work alone; it helped select which options went into paid media. Explore our client cases and estimate the impact using our ROI calculator at https://www.galaxies.com.br/pt/roi.

How much can you trust a decision tested with synthetic personas?

I hear this question more than any other, and it is fair. At research-industry events, I describe reactions to the concept in four stages: rejection (“it is just a guess”), understanding, application, and, after a few tests, enthusiasm. Moving from the first stage to the second requires transparency about how reliability is measured.

At Galaxies, we assess reliability by comparing persona responses with those of a control group of real people in the same segment, using questions that were not part of profile construction. Conversations are also evaluated continuously with the open DeepEval framework across five dimensions: relevance, bias, hallucination, fidelity to the profile, and toxicity.

Trust also means explaining the limits. Simulation is not formal market proof. When a board or regulator requires statistical validation, it serves as an initial filter for deciding what to take into the field. It does not predict sales volume; it compares options and explains preferences and barriers.

A persona is only as good as the data behind it. An audience absent from the client’s research should not be simulated from that research. Personas also answer within the domain they know; questions outside it should be redirected. Small percentage differences between runs are expected, so interpretation should focus on direction and reasons rather than tenths of a point.

Nexus or Simile: which simulation approach fits your question?

Anyone researching AI-based behavior simulation will soon encounter Simile, a US company founded by Stanford researchers connected to a 2025 study that created generative agents from two-hour interviews with 1,052 Americans [6]. Both companies work with simulation, but they address different questions. Simile starts with broad populations built from its own large-scale studies. Galaxies starts with each brand’s audience: Nexus personas are built from a client’s own real research, including Brazilian consumers.

If you need to understand how a country’s population might react to a policy change, broad-population simulation is better suited to the question. If you need to know how your own customers might react to a new message, or which of three ads should run for Black Friday in Brazil, a brand-specific audience is more relevant. The same distinction applies to internal employee questions versus testing a value proposition across your customer segments.

Where I see our differentiation is straightforward. The persona belongs to the brand, not an average population: it reflects the client’s research and answers “how would my customers respond?” Brazil is not a translation layered on afterward. Local payment habits, marketplace shopping, regional language, and category-specific skepticism are present when the underlying research was collected in Brazil.

There is also a human-led method around the tool. Nexus Agent is the natural-language entry point, with three study types: Persona Conversation, Questionnaire, and Stimulus Analysis. For enterprise accounts, a forward-deployed engineer works alongside the team, and each decision undergoes structured human review before reaching the committee. Client data is isolated under clear rules, with ISO 27001 certification and compliance with Brazil’s LGPD and the GDPR.

If the goal is to simulate whole populations, public policy, or employees across a multinational company, Simile’s proposition is broader than ours. For a consumer brand selling in Brazil and making decisions about its own audience, we believe our approach is the better fit.

Where does market research fit into this process?

It remains at the beginning and at the end. At the beginning, because Galaxies’ synthetic personas are built from real research conducted by the client or its partners, not generic internet data. At the end, because high-stakes decisions still warrant field validation.

In my talk, I used an image to explain the relationship. Traditional research provides a photograph: clear, costly, and true on the day it was taken. Today’s consumer behaves more like a film, changing scenes between the photo and the campaign launch. Companies can and should keep taking photographs. Without them, the film has nothing to work from. But the next campaign needs a frame from today, and simulation helps produce it.

In practice, research builds and refreshes the foundation. Simulation filters and compares options day to day. Research then returns to validate the few bets that reach the end of the decision funnel.

How can a marketing team start using decision intelligence?

You do not need to reorganize the department. Start with a decision already on the table and run it through the cycle. Choose one with a sign-off date in the next four weeks, preferably one involving media spend. Write a main hypothesis in a sentence the test could disprove. Bring at least two real alternatives, plus the current version when one exists.

Before testing, decide what will determine the choice: preference, barriers, clarity, or another measure. Test with audiences that reflect your actual customer base, not a generic consumer profile. Document the decision and compare it with the campaign’s real outcome 30 to 60 days later.

One recommendation comes from a mistake we made ourselves: do not test an asset that has already been approved just to confirm the choice. Once a decision is made, teams tend to interpret any result as validation, and the exercise loses its purpose.

Choosing has become marketing’s most expensive step

AI made production cheaper. What became expensive is choosing badly: a poor decision now reaches the market through more assets, in more channels, with less time to correct course. Decision intelligence is a practical response: treat the choice as a stage of work, with a hypothesis, alternatives, and a test, not as the final moment of a meeting.

I do not believe simulation will decide on behalf of the CMO. I believe it will change what reaches their desk. Instead of one confidently defended proposal, there will be two or three tested alternatives, with reasons and barriers by audience. The decision remains human. It is simply no longer made in the dark.

Bring a real decision to Nexus

Choose a messaging, creative, positioning, or audience decision you need to make in the coming weeks, and compare the options before spending. Explore how Nexus works.

Frequently asked questions about decision intelligence

What is decision intelligence in marketing?

It is the practice of structuring marketing decisions before execution: define the problem, form hypotheses, compare messaging, creative, pricing, or audience alternatives, and test how each performs. The goal is to choose based on evidence rather than conviction.

How does decision intelligence differ from business intelligence?

BI uses historical data to explain what has already happened. Decision intelligence acts before execution by comparing alternatives that have not yet reached the market. They complement each other: BI reveals the outcome, which informs the next decision.

Do synthetic personas replace market research?

No. At Galaxies, synthetic personas are built from a client’s or its partners’ real research. They help test and filter options before spending. Decisions requiring formal proof still go to the field, with fewer options and clearer hypotheses.

How much does it cost to test a decision with synthetic personas?

It depends on the scope and the number of decisions tested during the year. The useful comparison is between the cost of testing beforehand and the cost of a campaign that gets it wrong. Galaxies’ ROI calculator lets you estimate that with your own numbers.

How long does it take to test a campaign alternative?

A Nexus study runs in minutes. A structured verdict, including human review, is ready within 48 hours.

How do Galaxies and Simile differ?

Simile is a US company that simulates broad populations using its own research. Brazil-based Galaxies builds personas from each brand’s real research with Brazilian consumers. The former is better suited to questions about societies; the latter to a brand’s decisions about its own audience.

About the author

Daniel Victorino is CEO and co-founder of Galaxies, a consumer decision layer for marketing leaders at consumer businesses. Galaxies develops Nexus: Galaxies Lab, a platform for investigating and supporting decisions before spending, using synthetic personas built from real research with Brazilian consumers.

Sources

[1] Gartner, Hype Cycle for Artificial Intelligence 2025, definition and adoption estimate for decision intelligence, as reported by Aera Technology: https://www.aeratechnology.com/blogs/transformational-moment-decision-intelligence-2025-gartner-ai-hype-cycle/

[2] Clutch, “65% of Consumers Use AI to Research Products Before Making a Purchase,” January 22, 2026: https://www.businesswire.com/news/home/20260122526477/en/Clutch-Report-65-of-Consumers-Use-AI-to-Research-Products-Before-Making-a-Purchase

[3] PYMNTS Intelligence, Global Digital Shopping Index: Brazil Playbook, 2,273 consumers: https://www.pymnts.com/study/global-digital-shopping-index-new-brazilian-shopper/

[4] Makers and Uncover, O Futuro da Mensuração em Marketing, survey of 113 marketing leaders, May–June 2026: https://meet.uncover.co/o-futuro-da-mensuração-em-marketing-lkd

[5] Galaxies, client cases: https://www.galaxies.com.br/pt/clientes

[6] Stanford HAI, “Simulating Human Behavior with AI Agents,” May 2025: https://hai.stanford.edu/policy/simulating-human-behavior-with-ai-agents

We have partnered with Business Advantage, a UK-based B2B data and insight consultancy, to make Nexus audience simulation available to its clients, in a research territory defined by long buying cycles and audiences too small for any generic panel to reach.

In short: Galaxies announced a partnership with Business Advantage, a UK-based B2B data and insight consultancy founded in 1992. The partnership makes Nexus available to Business Advantage’s clients and combines the firm’s B2B research methodology with synthetic personas built from real research with an organization’s own audience. The goal is to test questions before fieldwork, not to replace it. Marketing Decision Intelligence is how we describe Galaxies’ work: helping marketing leaders decide on creative, claims, audiences and messaging before committing budget, using synthetic personas built from real research and human review.

On October 2, Daily Research News Online published the news of our partnership with Business Advantage, and on October 5 Research Live, the Market Research Society’s news outlet, also covered it. I want to explain here, with more context than a press note allows, why this partnership matters and for whom.

What is Business Advantage?

Business Advantage is a UK-based B2B data and insight consultancy, founded in 1992, offering market research, consulting and data services. It has worked for more than three decades across sectors including technology, CAD/CAM/PLM, printing, sustainability and telecommunications, combining multiple research methods with competitor and market analysis. That B2B research know-how, built over decades, is what the partnership brings together with Nexus technology.

Why is B2B research a different problem from B2C?

Consumer research deals with audiences of millions and relatively direct questions about preference, perception and purchase intent. B2B research faces a different reality: decision cycles that stretch for months, multiple decision-makers inside a single account, and expert audiences, such as technology buyers or category managers, too small for any generic panel to reach at sufficient volume and depth.

That is precisely the point Rachel Cope, Managing Director of Business Advantage, made when commenting on the partnership:

“B2B research brings unique challenges, and our partnership provides an opportunity to combine proven research methodologies with new technologies to answer questions in different and complementary ways.”

Nexus enters this picture by building synthetic personas from real research with an organization’s own audience, whether studies it already holds or data collected with research partners. For a specific B2B buyer who is hard to reach at volume, this means exploring hypotheses based on the company’s actual audience, not a generic panel that was never designed for that niche.

Does simulation replace fieldwork?

No, and that point is non-negotiable for us. This is what I said when commenting on the partnership for the trade press:

“The future of B2B intelligence is not about replacing research. It’s about giving organizations new ways to explore hypotheses, understand audience responses and inform decisions before they reach the market.”

In practice, research remains the foundation: it is what feeds the personas. Nexus works in the interval between one study and the next, to explore hypotheses, compare alternatives and help decide which questions deserve to go to the field. Talking to the real buyer remains irreplaceable. It is the same logic behind our synthetic persona use cases: the simulation layer exists to support the decision about what to research and how, never to eliminate research itself.

What does this partnership mean for Galaxies?

So far, the Nexus cases we have shared publicly center on consumer brands in banking, beauty, food and beverage, iGaming and technology. The partnership with Business Advantage takes the same Marketing Decision Intelligence approach into B2B, a territory that demands less volume and more precision: understanding what weighs on the decision of one specific corporate buyer, not a panel of thousands of consumers.

According to Research Live, the two companies said the partnership will make the tool available to Business Advantage clients.

For anyone leading research or insights at a B2B company, the invitation is simple: before budgeting the next fieldwork study, ask whether the hypotheses going into the questionnaire have been tested against any signal, or whether the questionnaire itself is the first time the question is being asked.

Closing: a partnership, not a shortcut

I like to close institutional pieces the way we closed the announcement: by reinforcing what the partnership is not. It is not a shortcut around fieldwork, nor a promise to replace decades of methodology with an AI model. It is the combination of two things that each solve only half the problem: the B2B research experience Business Advantage has built since 1992, and the Nexus simulation layer, built from each organization’s own real research.

Anyone who wants to see how simulation works in practice can explore Nexus or see the client cases already using the platform today.

Frequently asked questions

What is the partnership between Galaxies and Business Advantage?

A collaboration announced in October 2026 that combines the B2B research methodology of Business Advantage, a UK-based consultancy founded in 1992, with the Galaxies Nexus platform, allowing hypotheses about business audiences to be explored from real research with the organization’s own audience.

What is Business Advantage?

A UK-based B2B data and insight consultancy, founded in 1992, offering market research, consulting and data services across sectors including technology, CAD/CAM/PLM, printing, sustainability and telecommunications.

Who will have access to Nexus through the partnership?

According to Research Live, the two companies said the partnership will make Nexus available to Business Advantage clients.

Does audience simulation replace B2B market research?

No. Research is the foundation that feeds Nexus personas. Simulation helps explore hypotheses and compare alternatives between one study and the next, and fieldwork remains necessary to validate decisions with the real audience.

What data does Nexus use to simulate a B2B audience?

Real research with the organization’s own audience, whether studies it already holds or data collected with research partners. Personas are built from that data, not from a generic panel.

Sources cited

Research Live. Business Advantage and Galaxies work together on audience data, Oct. 5, 2026.

Daily Research News Online. Business Advantage Teams with Persona-Builder Galaxies, Oct. 2, 2026.

Business Advantage. business-advantage.com: research methodology, sectors served and company history.

On stage at the first edition of Makers Next, I explained why consumers now decide alongside an AI, and why yesterday’s research no longer describes the person your next campaign is about to meet.

In short: At Makers Next 2026, held on September 18 in São Paulo, Galaxies presented the talk “The Consumer Your Data Lost.” The thesis: consumers now decide alongside an AI, they change criteria faster than traditional research can track, and brands that shorten the gap between the consumer signal and the brand decision win the preference. Simulating a hypothesis before launch is one way to test without waiting weeks.

On the morning of September 18, I took the stage at Finca 50·62, in Morumbi, for the second talk of the day at Makers Next. It was 9:55 a.m., and Accenture Song had just opened the agenda talking about how AI is redesigning the way we work. The room was small by design: heads and managers from marketing, branding, CRM, media and creative. People who sign off on budgets.

I brought a question I’ve been asking clients for months, one that rarely gets a quick answer: when was the last time you actually sat down to get to know the consumer your brand is talking to today? Not the one from your last research wave. The one you have right now.

This piece lays out what I presented, with the data I used and where each figure comes from, plus what stuck from the conversations after the talk.

What was Makers Next 2026?

Makers Next is the marketing-leaders gathering created by Makers. The first edition took place on September 18, 2026, at Finca 50·62 in São Paulo, with 11 talks on the main stage and four Makers Lab sessions on a second stage capped at 30 people each. The organizers’ stated goal is to discuss what comes next with the people who actually have to decide.

The lineup covered AI in day-to-day operations, measurement and marketing mix modeling (Media Hero), creative effectiveness (Rastro), conversational marketing (Blip), CRM and retention (Braze), and branding (Droga5), among other topics. Our talk sat in the Research and the AI Consumer track.

One detail from the agenda stood out to me: the phrase “share of voice” came up twice that day. Once in our talk, applied to AI-generated answers. Later in the afternoon, in the title of Co-Creators’ talk, applied to the creator economy. I don’t know if that was intentional curation, but it’s a good sign that how brand presence gets measured is being rethought from more than one angle. The full agenda is on the Makers Next website.

Does the consumer decide alone, or alongside an AI?

I started with numbers that are already making the rounds. In January, Clutch published research showing that 65% of consumers use AI to research products before buying. In Brazil, the picture is even more pronounced: PYMNTS’ Global Digital Shopping Index, produced with Visa, surveyed 2,273 Brazilians in March and found that 53% had used AI during their most recent online purchase, ahead of the 46% figure for the United States.

The number that unsettled the room the most came from the same study. Between 2024 and 2026, ChatGPT’s share of product discovery in Brazil jumped from 2% to 32%, while Google’s share fell from 79% to 63%. In two years, a third of the discovery stage changed address.

“If you’re talking to one person, you’re only talking to half the decision.”

The other half happens somewhere the brand can’t see: inside an AI that searches, compares and recommends before the campaign even launches. And that AI isn’t just one system. Superlines analyzed 34,234 AI responses between January and February 2026 and found a 46-times difference in brand-citation rates between platforms. ChatGPT cited brands in 0.59% of responses; Perplexity, in 13.05%.

Translated into a marketing leader’s week: your brand doesn’t have a first impression. It has several, happening in parallel, each shaped by a different model. And brand strategy, at most companies, was built for just one.

The live test we ran in the room

At this point, I asked the audience to grab their phones, open ChatGPT, Gemini or Perplexity, and type: “What’s the best brand of [your category] in 2026?” Then to swap the question for “Which brand should I avoid in [your category]?”

“The room fell silent while the responses loaded. Some people found the brand to be well-positioned, while others did not. The AI was telling a story about each brand there, and no one from the brand was part of that conversation.”

I’d recommend running the same test before you keep reading. Note whether your brand showed up, in what position, and next to whom. It’s the cheapest diagnostic available right now.

What is AI Share of Voice, and why isn’t it in your report yet?

AI Share of Voice is how often a brand appears, gets cited, or gets recommended when someone asks an AI assistant a category question, measured against how often competitors do. It’s the AI-generated-answer equivalent of the metric marketing already tracks in media and search.

It matters because the click stopped being the default path. A SparkToro study using Similarweb panel data found that 68% of Google searches in the United States ended without a click in the first four months of 2026, up from 60% in 2024. When an AI Overview appears on the page, Seer Interactive measured a 61% drop in organic CTR. In that same study, brands cited inside the AI Overview earned 35% more organic clicks and 91% more paid clicks than brands not cited in the same search.

In other words: showing up inside the answer is now a condition for getting whatever click is still on offer. Tools to measure this already exist. What I see, talking to marketing teams, is that the metric rarely makes it into the monthly report. And the share lost through absence never shows up in a results meeting. It shows up later, on the balance sheet.

The first word about your brand is no longer yours.

How fast does your consumer data go stale?

Faster than your next campaign’s approval cycle. The same PYMNTS index shows Brazilians average 67 digital shopping days per month, the equivalent of 2.2 shopping activities a day, 8% more than in 2024. That’s a consumer who is learning the category’s vocabulary all the time.

Culture keeps the same pace. In the 2025 Sprout Social Index, 93% of consumers said it matters for brands to keep up with online culture, and 27% said jumping on a trend only works within the first 24 to 48 hours. At Galaxies we sum up the problem this way: a research cycle takes weeks; the conversation changes in days.

During the talk, I put a sector-specific question to the room:

• Beauty: skinification, skin cycling and barrier repair dominated the last few years. Which ones still matter, and what came after them?

• Pharma: mental-health OTC grew, the weight-loss category reorganized itself after the new injectables, longevity became a consumption habit. What has already fallen off that list?

• CPG: shrinkflation became a talking point, the GLP-1 effect reshaped buying patterns, private label gained share. Which of these is in your consumer’s vocabulary this week?

• Finance: installment Pix became routine, deepfakes became a real fear, betting got regulated. Which of these is the new conversation in your customer base?

I want to be clear on one point, because it tends to get misread. Traditional research hasn’t become useless. It’s a photograph: sharp, expensive, and true on the day it was taken. Today’s consumer is a film that keeps changing scenes between the photo and the campaign’s premiere. You can, and should, keep taking photographs, because without them the film doesn’t know what to shoot. What’s changed is that the next campaign’s decision needs a frame from today.

Historical data is your brand’s biography. It doesn’t describe the consumer you have now.

Journey or loop: how does the consumer decide now?

Since 2020, Google has called the messy stretch between first interest and purchase the “messy middle,” where consumers bounce between exploring and evaluating with no fixed order. Generative AI has made that middle even more circular. In the talk, I introduced what we call the AI Consumer Loop: the consumer discovers, understands, buys, learns, and comes back around. On every pass, they run into an AI that has already formed an opinion about the category, and often about your brand too.

The funnel is still useful for planning spend. As a description of how people actually decide, it falls short. Everyone in that room knows how to build brand awareness in the consumer’s mind. The question I left them with was a different one: does your brand have awareness inside the AI’s mind?

How many loops does the consumer run while your campaign waits for sign-off?

I proposed a simple calculation. Think about how many weeks passed between the brief for your last big campaign and the day it actually launched. Multiply that by seven. That’s how many days the consumer had to change their mind about your category while the company was still approving copy, creative, channel and budget.

On a campaign with eight weeks of approvals, that’s 56 days. If each week equals one lap of the loop, the message reaches someone who has already reshuffled their decision criteria eight times over, and much of that reshuffling happened inside an AI answering questions on the brand’s behalf.

This is where the money is made or lost. AI has sped up the consumer’s side. On the company’s side, the funnel still stalls at the human layer, when a piece of creative takes days to get signed off. Whoever shortens the gap between the consumer signal and the brand’s decision wins the consumer. And it all starts with the audience: creative, category, journey and measurement are all downstream of understanding who’s on the other side. If the read on the audience is out of date, everything after it inherits the error.

How do you test a hypothesis before it becomes a plan?

The page we published about the talk sums up the idea in one question: what if you could ask before you launch? That’s what we do with Nexus.

Nexus is Galaxies’ platform for simulating marketing decisions using synthetic personas built from a company’s own data. It doesn’t replace market research. It works as a simulation layer before launch, used to test messaging, creative, an offer or a positioning while there’s still time to change course. Research takes the photograph; simulation helps decide the next frame.

One concrete example: with Bradesco Seguros, each validation wave delivered 1,000 synthetic respondents in 48 hours, at a 93% lower cost per respondent (from R\$17.00 down to R\$1.20) and 10.5 times faster validation. The full case study is on our customers page. If you want to estimate the impact for your own scenario, the ROI calculator runs the numbers with your inputs.

The question I left in the room

I closed the talk with the same question that opens this piece. You know your brand better than anyone. When was the last time you stopped to get to know the consumer it’s talking to today?

Thank you to Makers for the invitation and for curating a first edition that put the people making these decisions in the same room. The page with the highlights from our talk is here. And if the question stuck with you too, see how Nexus works.

Frequently asked questions

What was Makers Next 2026?

It was the first edition of the marketing-leaders event organized by Makers, held on September 18, 2026 at Finca 50·62 in São Paulo. The event featured 11 talks on the main stage and four Makers Lab sessions, covering topics such as AI, measurement, branding, CRM and consumer research.

What was the topic of Galaxies’ talk at Makers Next?

The talk “The Consumer Your Data Lost,” presented by Daniel Victorino, CEO of Galaxies, covered how AI has become part of purchase decisions, why consumer data goes stale quickly, and how to shorten the gap between the consumer signal and the brand’s decision.

What is AI Share of Voice?

AI Share of Voice is how often a brand appears, gets cited, or gets recommended when someone asks an AI assistant, such as ChatGPT, Gemini or Perplexity, a category question, compared with how often competitors do.

Do synthetic personas replace market research?

No. Traditional research remains the foundation for understanding the consumer. Synthetic personas, like the ones in Nexus, work as a complementary simulation layer, used to test marketing hypotheses before launch, while there’s still time to change the plan.

Why does consumer data go stale so fast?

Because consumers interact with a category constantly. In Brazil, the average is 67 digital shopping days a month, according to PYMNTS, and a growing share of product research now runs through AI assistants. A research cycle that takes weeks can land after the consumer has already changed their criteria.

Sources cited

• Makers. Makers Next 2026. Event agenda and details.

• Clutch. Agentic Commerce: How AI Is Changing Online Shopping Behavior, Jan. 2026.

• PYMNTS Intelligence and Visa Acceptance Solutions. Global Digital Shopping Index: Brazil Playbook, 2026.

• Payments Industry Intelligence. Brazil’s AI shoppers point to the next phase of Agentic Commerce, Aug. 2026.

• Superlines. AI Search Statistics 2026, Mar. 2026.

• SparkToro. In 2026, Less than One Third of Google Searches Still Send a Click, Jun. 2026.

• Sprout Social. The 2025 Sprout Social Index, Jan. 2025.

Over the next 18 months, I expect decision engines to stop being a competitive advantage and become a basic expectation of the CMO role, just as BI dashboards are today. I make this forecast aware that trend articles written by technology CEOs can be vague enough never to be wrong. I would rather risk being wrong with a specific prediction.

Why I Believe This Forecast Is Right?

The Brazilian CMO X-Ray showed a role whose accountability changed faster than its tools: it is now measured by pipeline and ROI, yet largely formed through a brand-focused repertoire. On the other side, marketing AI adoption data shows high investment intent and still-low maturity. These tensions will not resolve themselves.

My view is that we will see the same cycle as other tool categories: what is now used by teams ahead of the curve becomes a minimum standard quickly, because teams that do not adopt visibly become slower than their competition.

Three Likely Changes Through 2028

1. Simulation becomes routine, not a special project. Just as no CMO approves a campaign today without checking a dashboard, the trend is that no relevant decision will be approved without first being simulated with the audience.

2. The expectation for tested decisions increases. As decision engines become accessible, not testing before investing stops being accepted as a lack-of-time issue and becomes a lack-of-rigor issue.

3. The CMO’s valued repertoire changes. Creativity and brand remain essential, but they will be evaluated alongside data literacy and tested decision-making — complementary rather than isolated capabilities.

Where Could I Be Wrong?

The biggest risk in this forecast is pace, not direction. I have seen marketing-technology cycles where adoption expected in two years took five, usually because organizational behavior changes more slowly than available tools. If I am wrong, it will likely be through calendar optimism.

What Does This Require from Today’s Decision-Makers?

Teams that begin incorporating simulation into their approval routine now will reach 2027 with an advantage that is not only technological but habitual. The learning curve is easier to travel gradually than all at once under pressure. To begin, understand how a marketing decision engine works.

Frequently Asked Questions

How will AI change the CMO’s work by 2027?

This article’s forecast is that simulating marketing decisions before investing will stop being a differentiator and become a basic expectation of the role, just as performance dashboards are today.

Will brand and creative skills lose importance?

No. They will be evaluated alongside data literacy and tested decision-making. The capabilities become complementary, not replacements for one another.

What can a marketing team do now to prepare?

Incorporate audience simulation into the approval routine, starting with a specific bottleneck rather than waiting for the practice to become mandatory under market pressure.

Arrive First at the Routine Everyone Will Adopt

Every change in market expectation passes through a phase where it is a differentiator before becoming a minimum standard. Simulating marketing decisions before investing is in that transition now. Companies treating it as a routine today are not only testing a technology; they are building the habit the market will soon require.

Prepare your operation for 2027. Schedule an executive demo.

Synthetic personas simulate behavioral patterns with high accuracy, but they are not infallible. Like any generative AI system, they are subject to bias and responses that can drift away from their underlying data, the equivalent of what language models call hallucination. Galaxies evaluates every persona with the DeepEval framework across five dimensions: relevance, bias, hallucination, faithfulness and toxicity, before considering a simulation reliable enough for decision-making.

Why Is Talking About Limits Part of Credibility?

The most common objection from an enterprise CMO facing any generative AI tool is simple: is this reliable? Ignoring the question, or responding only with success statistics, tends to generate more distrust than security. After years building AI systems, I have learned that the most technically sophisticated client is precisely the one most skeptical of anyone promising perfection.

Being direct about where a technology has limits is not a positioning weakness. It is what separates a serious decision tool from a marketing promise. This article addresses what synthetic personas do well and where additional validation remains necessary.

How Does Galaxies Evaluate Each Persona’s Quality?

Every persona built in Nexus goes through a structured evaluation with DeepEval, an open-source framework that measures the quality of language-model responses across five dimensions.

Relevance. Whether the response actually addresses the question asked rather than drifting into generalities.

Bias. Whether responses display differential treatment based on demographic characteristics of the persona.

Hallucination. Whether the response contains information not documented in the data that grounds that persona.

Faithfulness. Whether there are contradictions between the response and the persona’s source data.

Toxicity. Whether the response contains offensive, hostile or inappropriate language.

This process makes it possible to treat a simulated response as reliable for informing a decision. Confidence is supported not by the absence of error, but by a structured process for identifying it before it reaches the final decision.

A Concrete Example of a Faithfulness Failure

Imagine a persona built from data describing an audience that values low price above all else. If that persona, in a conversation about a new premium line, responds enthusiastically about paying more for exclusivity, it is a signal of low faithfulness: the response contradicts the source data underlying that profile. This is exactly the kind of inconsistency the evaluation is designed to catch before it reaches the marketing team as a trustworthy insight.

Where Do Synthetic Personas Work Well?

They help test messaging, creative, price and positioning hypotheses with a speed and depth conventional processes cannot sustain; explore multiple profile reactions in parallel; and simulate low- to medium-risk decisions where a clear direction is enough to decide with confidence. For a practical application, see how to validate campaign creatives with AI.

Where Is Additional Validation Still Recommended?

For decisions with very high regulatory or financial risk, where the cost of an error justifies an extra confirmation layer. It is also recommended in highly specific cultural or behavioral contexts that are underrepresented in the data grounding a persona, and in high-investment final decisions where a complementary check with real people reduces remaining risk further.

Frequently Asked Questions

Do synthetic personas have limitations?

Yes. Like any generative AI system, they are subject to bias and responses that drift from their grounding data. Galaxies evaluates these dimensions with DeepEval before considering a simulation reliable.

How does Galaxies avoid hallucinations in persona responses?

Each persona is evaluated in five DeepEval dimensions: relevance, bias, hallucination, faithfulness and toxicity. The process compares the generated response with the documented data that grounds that persona.

Do synthetic personas replace every form of human validation?

Not for decisions with very high regulatory or financial risk, or in contexts poorly represented by available data. In these cases, an additional confirmation layer with real audiences remains recommended.

Trust Is Built Through Transparency, Not Silence

No simulation technology is perfect, and pretending otherwise does not build trust. It builds distrust when the limit appears in practice. Being clear about where synthetic personas work well and where additional validation still matters is what supports responsible use of this technology in decisions that truly matter to the business.

Talk to our team about the ideal study design for your case. Schedule a technical session.

Analyzing marketing data with AI does not mean accumulating more dashboards. It means combining declared data, what the audience says, with simulated data, how the audience may react, to reach a specific decision. From a technical perspective, the most common architecture mistake I see is stacking tools that describe the past while having none that project the future of a decision not yet made.

The Problem of Full Dashboards and Empty Decisions

Many marketing operations now have more data than they can use: site analytics, CRM, social media and paid media. When reviewing client data stacks during implementation, the pattern repeats: the problem is rarely missing raw data, but difficulty turning volume into a specific decision about the next campaign or launch.

Historical data is excellent for explaining what has already happened, but it does not answer the most important planning question on its own: what would happen if the company made a different decision from the one it made in the past? No dashboard, however well built, answers a counterfactual question about a decision that is not yet in the data.

Declared Data versus Simulated Behavior

Declared data, what an audience says in an interaction, evaluation or form, is valuable but limited: people do not always behave as they say they would. Simulated behavior data, generated from personas built using a company’s real data, complements this limitation by projecting how different profiles may react to a specific hypothesis before it happens.

Neither replaces the other. Combining what the audience says with how personas react to a simulated scenario gives the team a more complete decision foundation than either layer alone. To understand how that foundation is built, read how synthetic personas are created.

How to Combine These Data Sources in Practice

1. Identify the decision historical data cannot answer. It is usually a future-facing question: which message would work better, what price would the audience accept or which segment deserves more investment.

2. Build personas from the data the company already has. CRM, proprietary research and behavioral data feed a more precise audience portrait.

3. Simulate the specific hypothesis with those personas. Test the open decision directly instead of inferring an answer from historical data that does not cover that exact scenario.

4. Decide with both layers combined. History shows the trend; simulation tests the specific hypothesis it does not cover. Explore the marketing decision engine.

A Technical Caveat Worth Explaining

One mistake I have seen teams make is treating simulation results with the same statistical weight as historical data with thousands of real observations. They are not the same. Historical data describes what happened, with the normal error margin of any measurement; simulation projects probable behavior, with a different uncertainty that depends on the quality of data used to build the persona.

My recommendation is to treat simulation results as qualified hypotheses to test on a smaller scale before committing an entire budget, not as certainty equivalent to consolidated historical performance.

Frequently Asked Questions

How can AI be used to analyze marketing data?

Combine historical data, what has happened, with audience simulation based on synthetic personas, what could happen with a decision not yet made, rather than relying solely on a retrospective dashboard.

What is the difference between declared data and simulated behavior?

Declared data is what an audience says about itself. Simulated behavior is the projected response of personas built from real data to a specific scenario. They complement each other but carry different levels of uncertainty.

Do I need a large data team to combine these sources?

Not necessarily. Nexus is designed so that marketing teams can build personas from existing company data and run simulations without relying on a dedicated technical team for every test.

From Raw Data to Decision

The value of a marketing-data operation is not the volume of information accumulated, but the ability to turn that volume into a specific decision about what to do next. Combining historical data with simulation, while respecting each layer’s weight and uncertainty, closes the gap between having data and having a decision.

Turn dispersed data into decisions. Explore Nexus.