The Data Team Is Losing Budget
Gartner expects worldwide software spending to reach $1.43 trillion in 2026, yet the competitive intelligence teams getting the most executive attention are hiring people trained to watch buyers, not people trained only to wire dashboards (Gartner IT spending forecast, February 2026). That’s the counterintuitive move inside enterprise go-to-market organizations: the next scarce CI hire isn’t another data engineer, it’s an ethnographer who can sit inside buying committees, sales calls, partner channels, procurement reviews, and user communities, then explain why the numbers moved before the dashboard can prove it.
The reason is simple: most CI data stacks now see the same public signals. Product pages, job posts, pricing screenshots, customer reviews, SEO traffic, social chatter, analyst notes, sales-call snippets, and patent databases are easier to scrape and summarize than they were three years ago. That made business intelligence hiring more efficient, but it also compressed advantage. When every vendor can point an AI crawler at the same sources, the edge shifts toward interpreting messy behavior: why a CIO says security is the deciding factor but awards the deal to the vendor with the lower migration risk, why a procurement team delays renewal after a competitor’s roadmap leak, or why a business unit keeps using a shadow tool after headquarters standardizes on a suite.
In 2026, the winning CI team is less like a data factory and more like an intelligence desk with field researchers attached. The best teams still need clean data pipelines, CRM joins, call-intelligence feeds, and alerting. They just don’t treat those assets as the strategy. The strategy is converting qualitative market research into sharper decisions on pricing, packaging, product roadmap, partner defense, and sales plays. That’s why ethnography in B2B research has moved from design-research departments into competitive intelligence talent plans.
The weak research context available for this topic shows the core problem: raw web artifacts can be noisy, incomplete, or irrelevant. A skipped search result and a stray ID aren’t intelligence. They’re evidence that automated collection without human interpretation can create the illusion of coverage. MarketIntel’s own work on enterprise technology markets points to the same operating truth: buyers don’t make strategic purchases as clean data points. They make them in rooms shaped by fear, incentives, politics, sunk cost, compliance risk, and career risk.
$116 Billion Chases Better Judgment
The market research services market is expected to reach $96.77 billion in 2026 and $116.02 billion by 2030, a 4.6% CAGR from 2025 to 2030, according to The Business Research Company’s 2026 market research services report (The Business Research Company, 2026). That’s the broad services TAM for research work, including marketing research, public opinion research, qualitative research, quantitative research, and mixed-method work. The more relevant serviceable market for CI teams is smaller: analyst subscriptions, digital intelligence platforms, win-loss research, expert interviews, market monitoring, and custom buyer research. Based on vendor revenues and published market boundaries, this CI-adjacent SAM is roughly $18 billion to $25 billion in 2026, an analyst estimate using Gartner, Forrester, Similarweb, AlphaSense, Ipsos, and market research services disclosures.
The software layer is growing faster than the research-services base. Gartner estimated the data and analytics software market at $175.17 billion in 2024 after 13.9% growth, with data science and AI platforms up 38.6% and nonrelational database management systems up 22.7% (Gartner market share research, 2025). Gartner’s later opportunity map put data and analytics software at $175 billion in 2025 and forecast $358 billion by 2029, implying a 15.4% CAGR from 2025 to 2029 (Gartner market opportunity map, 2025). IDC’s business intelligence and analytics software forecast also frames GenAI, embedded analytics, and self-service analytics as growth drivers through 2028 (IDC BIA forecast, 2024).
Those figures explain the inflection. From 2018 to 2022, CI budgets mostly followed the analytics stack: more warehouse spend, more dashboards, more intent data, more automated alerts. From 2023 to 2025, generative AI lowered the cost of summarizing public and internal text, which made basic monitoring cheaper. By August 2026, the scarcity is no longer collection; it’s interpretation under uncertainty.
Regional differences matter. North America remained the largest market for market research services in 2025, while Western Europe was cited as the fastest-growing region in the 2026 report (The Business Research Company, 2026). Europe is also more compliance-sensitive because the EU AI Act’s enforcement powers for general-purpose AI obligations begin on 2 August 2026 (European Commission AI Act Service Desk, 2026). Asia-Pacific demand is more uneven: large technology, manufacturing, and financial groups are funding CI, but many mid-market firms still buy project-based research rather than permanent intelligence teams, an analyst estimate based on vendor disclosures and regional revenue splits.
The current shift is a budget reallocation, not a rejection of data engineering. A 2026 enterprise CI operating model typically still needs a data owner for source governance, a product marketer for sales enablement, and an analyst for market sizing. The new hire is the person who can validate whether the data is describing the buyer’s real decision process. That’s why qualitative market research is moving closer to revenue operations, product strategy, and M&A diligence.
The Platforms Rewriting The Workbench
AlphaSense has become the reference platform for AI-powered market intelligence because it combines enterprise search, filings, broker research, earnings transcripts, expert-call content, and GenAI workflows. The company said it surpassed $400 million in annual recurring revenue in March 2025, up from $200 million in April 2024, and served more than 6,000 customers, including 88% of the S&P 100 (AlphaSense company release, 2025). Its defining move was the $930 million acquisition of Tegus in June 2024, which added more than 150,000 expert interview transcripts covering more than 35,000 public and private companies, giving CI teams a much richer base of human-source evidence.
Gartner remains the dominant C-suite advisory franchise, and its own numbers show why incumbents still matter. Gartner reported 2025 revenue of $6.5 billion, with $5.2 billion of contract value and $3.9 billion of Global Technology Sales contract value (Gartner filings and results, 2026). In April 2026, Gartner published its first Magic Quadrant for Competitive and Market Intelligence Platforms, naming vendors such as AlphaSense, Crayon, Klue, Evalueserve, Market Logic, Stravito, and Valona Intelligence; that formalized the category for procurement teams that previously treated CI tools as sales enablement add-ons (Gartner Magic Quadrant abstract, 2026).
Forrester is repositioning around buyer insight and AI-assisted advisory access while dealing with a weaker financial base. The company reported 2025 revenue of $396.9 million, down from $432.5 million in 2024, and contract value of $292.4 million, down 6% year over year (Forrester results, 2026). Its Forrester Market Insights service emphasizes buyer knowledge, competitive positioning, market forecasts, and access to research grounded in annual surveys of more than 675,000 consumers, business leaders, and technology leaders, which gives it a defensible position when clients want advisory interpretation rather than just another dashboard (Forrester product materials, 2026).
Similarweb is pushing competitive intelligence toward proprietary digital exhaust: traffic, app, search, advertising, product, and company-level signals. The company reported 2025 revenue of $282.6 million, up 13%, with 6,128 customers and 454 customers above $100,000 ARR as of 31 December 2025 (Similarweb results, 2026). Its 2026 move was commercial proof around AI data: Similarweb announced in June 2026 that ARR surpassed $300 million after two multi-year enterprise contracts with roughly $47 million in total contract value, while its GenAI intelligence product had approached 200 customers and about $3 million of ARR after launching in Q3 2025 (Similarweb releases and shareholder letter, 2026).
Klue sits closer to revenue teams than most research platforms. It’s a private competitive enablement company, so revenue isn’t disclosed, but it reported more than 250,000 users globally as of 2025 and had raised a $62 million Series B led by Tiger Global in 2021 (Klue company materials and funding announcement). The late-2025 move was acquisitive: Klue bought Goldpan.ai in March 2025 and Ignition in September 2025, then launched Compete Agent, Auto Insights, and win-loss AI workflows aimed at turning CRM, Gong, win-loss interviews, internal documents, and public sources into deal-specific intelligence.
Crayon is the pure-play competitor most associated with battlecards, competitor monitoring, and content delivery into sales workflows. The company is private and doesn’t publish revenue, but it had raised roughly $38 million in Series B funding in 2021, and its 2025 product move was the launch of competitive intelligence APIs and an MCP server designed to feed Crayon content into ChatGPT, Glean, Microsoft Copilot, Google Gemini, Slack, Teams, and custom assistants (Crayon product release, 2025). That move shows where the category is going: CI platforms are becoming knowledge infrastructure for enterprise AI agents, not just alert inboxes for product marketers.
The share gainers are the vendors with proprietary human context or proprietary behavioral data. AlphaSense gains from expert transcripts and financial workflows, Similarweb gains from digital activity data, Klue gains from deal-level win-loss context, and Gartner keeps power where board-level validation matters. Generic data aggregation loses value because AI makes aggregation cheap.
The Rule That Changed The Org Chart
The specific trigger is the EU AI Act enforcement timeline, especially the 2 August 2026 start of enforcement powers for prohibited AI practices, transparency requirements for certain AI systems, and rules for general-purpose AI models (European Commission AI Act Service Desk, 2026). The Act doesn’t say companies must hire ethnographers. It changes the operating environment in a way that makes purely technical CI less sufficient: vendors now need clearer evidence about model limitations, downstream use, buyer risk tolerance, human oversight, and how AI claims are understood in market-facing workflows.
The connection is direct. The European Commission’s guidelines state that general-purpose AI model providers must maintain technical documentation, provide downstream providers with information about capabilities and limits, apply copyright policies, and publish training-content summaries for models placed on the market after 2 August 2025, with enforcement powers beginning on 2 August 2026 (European Commission GPAI guidelines, 2026). For enterprise buyers, that turns AI messaging into a procurement risk question. For vendors, it turns product positioning into a regulated claim surface.
NIST’s AI Risk Management Framework gives the organizational reason ethnographers are showing up in CI team structure. NIST says trustworthy AI requires context of use and that interdisciplinary actors, including domain experts, socio-cultural analysts, human factors experts, governance experts, product managers, evaluators, and members of impacted communities, should help establish context (NIST AI RMF, 2023). NIST also warns that converting social and individual decision practices into measurable quantities can strip away needed context, which is exactly the failure mode in dashboard-led CI.
By August 2026, the macro shift is not just AI adoption; it’s AI accountability entering procurement. A data engineer can identify that a competitor changed its model card, pricing page, or security documentation. An ethnographer can explain whether the buyer believes the change, whether legal treats it as material, and whether sales teams can turn it into a defensible competitive wedge.
Three Risks Few Boards Price
The first risk is false confidence from automated monitoring, with a 55% probability across enterprise CI programs over the next 12 months, an analyst estimate. The mechanism is straightforward: AI systems collect more weak signals than human teams can validate, then summarize them into clean narratives. Affected players include CI platform vendors, revenue operations leaders, and product marketers using generic copilots for battlecards. The likely timeline is immediate through mid-2027, as companies connect CI repositories to internal AI assistants and assume the answer quality is higher than it is. The damage is not just a wrong fact; it’s a sales team repeating a weak claim in a competitive deal.
The second risk is compliance drag, with a 40% probability for AI-heavy vendors selling into Europe by late 2026, an analyst estimate grounded in the EU AI Act enforcement schedule. The mechanism is slower review of claims about model performance, data sources, and workflow automation. Affected players include AlphaSense, Similarweb, Klue, Crayon, enterprise AI search vendors, and any B2B software company embedding third-party models into market intelligence workflows. The timeline runs from August 2026 to August 2027 as legal, security, and procurement teams decide how much proof they need before approving AI-enabled CI tools.
The third risk is talent misallocation, with a 50% probability for companies that respond to AI by over-hiring prompt engineers and under-hiring field researchers, an analyst estimate. The mechanism is budget optics: technical hires are easier to justify because they map to systems, while qualitative researchers can look like discretionary advisory spend. Affected players are mid-market SaaS firms, private-equity portfolio companies, and industrial technology vendors entering new segments. The timeline is the 2027 planning cycle, when executives will discover that more automation hasn’t improved win rates against better-positioned competitors.
The tail risk is ethnographic capture. A small number of vivid buyer interviews can overweight executive judgment if the research design is sloppy. This is underpriced because qualitative methods are being adopted by teams that may not know sampling discipline, moderator bias, or how to triangulate interview evidence against pipeline and market data. The probability is lower, roughly 20% over 18 months, but the impact is high in M&A, category-entry decisions, and pricing resets.
Enterprise buyers
Enterprise buyers should stop treating CI as a content function and fund it as a decision function. The first action is to require every major competitive claim to carry evidence grading: observed buyer quote, verified document, sales-call pattern, third-party data, or analyst estimate. The second is to attach CI researchers to three moments that move revenue: late-stage losses, renewal risk, and product-roadmap prioritization. The third is to separate signal collection from interpretation. A platform such as Similarweb can show traffic share shifts across 100 million-plus websites and 4 million-plus apps (Similarweb product materials, 2026), but a trained field researcher has to explain why those shifts matter to a buyer making a budget decision.
Investors
Investors should diligence CI maturity the same way they diligence sales productivity. In growth equity and PE deals, the practical test is whether management can name the top three reasons buyers switch to competitors, backed by win-loss evidence rather than sales anecdotes. Investors should also track qualitative research capacity inside portfolio companies because it affects pricing power, churn defense, and cross-sell planning. A software company with $50 million ARR that misunderstands why it loses regulated buyers can burn several million dollars a year in wasted product and sales spend, an analyst estimate based on typical enterprise SaaS go-to-market cost structures.
Vendors
Vendors should build products that make human judgment auditable. That means source provenance, confidence levels, interview tagging, CRM linkage, and a clear path from field evidence to sales action. Klue’s acquisition of Ignition and launch of Compete Agent point toward AI-assisted workflows, while AlphaSense’s Tegus acquisition shows the value of curated expert interviews. The winning product pattern is not replacing researchers. It’s making the researcher’s evidence easier to search, compare, challenge, and reuse across sales, product, and executive teams.
The Next Two Budget Cycles
The base case has a 60% probability: by the end of 2027, large B2B vendors will rebalance CI team structure toward mixed teams with one qualitative lead for every two to four analytics or enablement roles, an analyst estimate. Software spend keeps rising, but the marginal dollar shifts to interpretation, win-loss research, expert networks, and buyer ethnography. Gartner’s 2026 forecast of $1.43 trillion in software spending and The Business Research Company’s $96.77 billion market research services estimate both support the direction: more digital spend creates more ambiguity, not less (Gartner, 2026; The Business Research Company, 2026).
The contrarian view has a 25% probability: AI agents improve enough that ethnographic hiring stays niche. In this scenario, platforms such as AlphaSense, Similarweb, Klue, and Crayon ingest more private content, connect to CRM and call systems, and generate sufficiently accurate deal guidance for most teams. The result would be fewer dedicated ethnographers and more hybrid product marketers trained in qualitative methods. This view is plausible, but it assumes companies can solve source trust, buyer context, and organizational politics through software alone.
The downside scenario has a 15% probability: budget pressure freezes CI hiring and pushes companies back to cheaper automated summaries. Forrester’s 2025 revenue decline to $396.9 million and guidance for lower 2026 revenue show that advisory and research budgets aren’t immune to cuts (Forrester filings, 2026). If CFOs treat qualitative research as discretionary, teams may defer hiring until competitive losses become visible in churn or win rates.
The leading indicators are concrete. Watch the share of CI job postings that mention win-loss interviews, ethnography, buyer research, or qualitative methods; watch the number of CI platforms adding interview repositories and source-confidence scoring; watch procurement language around AI explainability and model-risk documentation after 2 August 2026. If those three indicators rise together, the ethnographer thesis is working.
Seven Takeaways For Skimmers
- Gartner’s $1.43 trillion 2026 software spending forecast makes CI more important because buyers face more vendor claims, not fewer.
- The market research services market is $96.77 billion in 2026, but the CI-adjacent serviceable market is roughly $18 billion to $25 billion, an analyst estimate.
- AlphaSense’s $400 million-plus ARR and Tegus acquisition show that expert interviews and private content are becoming core CI assets.
- Similarweb’s $300 million-plus ARR in 2026 shows that proprietary behavioral data is still valuable when public web data becomes easier In short,.
- The EU AI Act’s 2 August 2026 enforcement timeline turns AI product claims into procurement and competitive-risk evidence.
- Ethnographers help CI teams explain buyer behavior that dashboards can detect only after the deal is already lost.
- The best 2027 CI teams will combine automated monitoring, win-loss research, field interviews, and source-graded evidence inside sales and product workflows.
How should a CFO measure the return on hiring an ethnographer for CI?
A CFO should tie the role to three measurable pools: competitive win rate, discounting, and churn prevention. The cleanest test is a 90-day pilot focused on one segment, one competitor set, and one revenue motion, such as enterprise renewals above $250,000 ARR, an analyst estimate for a typical SaaS threshold. The ethnographer should produce coded win-loss themes, buyer-risk maps, and sales plays that can be compared against CRM outcomes. Klue’s win-loss positioning and AlphaSense’s expert transcript model show the same financial logic from different angles: human-source evidence becomes valuable when it changes seller behavior or product prioritization. The CFO shouldn’t fund open-ended research theater. The CFO should fund a decision pipeline with a baseline, an intervention, and a measured change in win rate or retained ARR.
Does this mean data engineers no longer belong in CI teams?
Data engineers still belong in CI teams, but their role is shifting from center stage to infrastructure. CI still needs clean ingestion from CRM, call intelligence, product telemetry, review sites, pricing trackers, web monitoring, and analyst repositories. Gartner’s data and analytics software market estimate of $175.17 billion in 2024 shows that the enterprise stack isn’t shrinking (Gartner market share research, 2025). The issue is that data engineering alone can’t explain why a buyer ignores a feature gap, why a partner steers deals toward a rival, or why a renewal committee uses compliance language to mask budget pressure. The better model is paired work: data engineers establish trusted signal flow, while ethnographers validate meaning through interviews, observation, and buyer journey analysis.
Which vendor categories benefit most from ethnography-led CI?
The biggest beneficiaries are categories where buying committees are large, switching costs are high, and public product data understates the real decision. Enterprise software, cybersecurity, industrial automation, cloud infrastructure, healthcare technology, and financial data platforms fit that profile. Similarweb can quantify digital attention, AlphaSense can organize financial and expert evidence, and Gartner can validate market categories for boards, but ethnographic work explains how committees interpret those inputs. In cybersecurity, for example, a buyer may say detection quality matters most but choose the vendor with better incident-response confidence. In cloud infrastructure, procurement may cite price while the architecture team fears lock-in. Those hidden motives shape actual share movement, which makes qualitative market research a competitive asset.
What should a PE investor ask during commercial diligence?
A PE investor should ask management to show the last ten competitive losses, the buyer-stated reason for each loss, the internal hypothesis, and the evidence used to reconcile the two. If the company can’t answer, the CI function is likely a reporting function, not a learning system. The investor should also ask whether sales calls, churn interviews, lost-deal interviews, analyst feedback, and competitive product changes are connected in one operating rhythm. Forrester’s 2025 revenue decline and Gartner’s $6.5 billion 2025 revenue show that even research incumbents face pressure and category shifts (Forrester and Gartner filings, 2026). Commercial diligence should test whether the target has a repeatable way to learn from buyers before competitors expose the weakness in pricing, onboarding, or roadmap credibility.
How should a CTO evaluate AI-enabled CI tools in 2026?
A CTO should evaluate source provenance, permission controls, integration depth, auditability, and model-risk controls before evaluating the quality of generated summaries. The EU AI Act’s 2 August 2026 enforcement start for GPAI obligations raises the cost of vague AI claims, especially for vendors selling into Europe (European Commission, 2026). Crayon’s MCP server and Klue’s MCP server show that CI content is moving into AI agents, which makes governance more important. The CTO should ask whether the tool can distinguish a verified buyer interview from a public blog post, whether it can restrict sensitive win-loss evidence, and whether it logs how an answer was produced. Without that evidence trail, AI-enabled CI becomes a compliance and sales-risk surface.
The Human Edge In Machine Markets
Competitive intelligence talent is being repriced because AI has changed the economics of information. Collection is cheaper, summarization is faster, and public data coverage is broader. That should have made data-heavy teams stronger. Instead, it exposed the harder problem: companies don’t lose strategic deals because they missed one more web alert. They lose because they misunderstand what buyers fear, what committees reward, and what competitors are making easier at the exact moment procurement is ready to act.
The practical move for 2026 is to build CI around evidence quality. Data engineers keep the pipes clean. Product marketers turn insight into field action. Ethnographers uncover the buyer logic that doesn’t show up in structured fields. Executives should fund the combination, then demand proof that the research changes pricing, roadmap tradeoffs, renewal saves, and competitive win rates.
The vendors gaining ground will be the ones that blend proprietary data with human context: AlphaSense through expert content, Similarweb through digital behavior, Klue through deal workflows, Crayon through embedded compete content, and Gartner through executive validation. The teams gaining ground will mirror that pattern internally. By 31 December 2027, at least one-third of enterprise B2B CI job descriptions for director-level roles will explicitly require win-loss interviewing, ethnographic research, or qualitative buyer-research experience, an analyst estimate that can be tested against public job postings.