Quick answer
The measure most intelligence functions report is the one least likely to protect their budget. Programs early in their maturity count outputs: briefs produced, battlecards maintained, alerts sent, reports delivered. Mature programs count decisions influenced, win-rate impact on competitive deals, and estimated revenue effect. That gap matters in 2026 because finance leaders are making selective cuts rather than blanket ones, which puts the burden on each function to show it is essential. The most practical step is also the least sophisticated: keep a decisions-influenced register, a simple log of decisions your analysis shaped, and review it quarterly with your executive sponsor.
Key takeaways
- Gartner found 64% of CFOs are planning overhead to grow more slowly than revenue in 2026, with cuts aimed at specific functions rather than applied across the board.
- Output volume is the easiest measure to produce and the easiest to cut, because it reads as cost rather than contribution.
- In Segment8's 2026 survey, the share of programs tracking a register of decisions influenced rises from 12% at stage 2 maturity to 76% at stage 4.
- Most intelligence functions never measure their value at all, and the usual reason is that nobody senior asked.
- The programs with the strongest executive engagement are indispensable to one specific recurring decision.
- A decisions-influenced register costs nothing, takes a spreadsheet, and can be backfilled from the last two quarters.
What are CFOs actually cutting in 2026?
Gartner surveyed 142 CFOs and senior finance leaders about their 2026 budget assumptions, fielding the work between August and September 2025. Sixty-four percent said they are planning for selling, general and administrative costs to grow more slowly than revenue. Fifty-four percent expect that gap to be one to five percentage points.
The more useful detail is how they intend to get there. In Gartner's words, rather than implementing broad, across-the-board cuts, CFOs are "calibrating their approach, seeking incremental cost containment that preserves essential capabilities."
Selective is better news than blanket, and harder to survive. A uniform five percent cut asks nothing of you. A selective cut asks you to demonstrate that you are one of the essential capabilities, in terms a finance leader recognizes.
The functions most often named for reduction were human resources at 57%, corporate IT at 53%, legal and compliance at 40%, corporate finance at 36%, and marketing at 27%. Market research, insights and competitive intelligence teams usually sit inside one of those cost centers rather than carrying their own line, which means they are frequently cut by proxy, as part of a reduction aimed at the department around them.
Why does counting outputs fail as a measure?
Because it describes effort, not effect.
A number of reports delivered tells a CFO how busy the team was. It does not tell them what would have gone differently without the team. Worse, a high output count read in a cost-reduction conversation can work against you: it looks like volume that could be trimmed without anyone noticing.
Avner Barnea, writing in Competitive Intelligence Magazine, puts the problem precisely. When measurement happens at all, he observes, "it usually focuses on the value of CI products, rather than how those outputs are used to influence decisions."
That distinction is the whole argument. The product is the report. The value is the decision the report changed.
What do mature intelligence programs measure instead?
Segment8 surveyed 612 competitive intelligence practitioners across 41 countries between July 2025 and March 2026, and broke measurement practice down by how mature each program is. The pattern is consistent across every metric.
At stage 2 of their maturity model, 92% of programs measure the volume of outputs delivered. At stage 4, that falls to 61%. What rises in its place:
- A tracked register of decisions influenced: 12% at stage 2, 76% at stage 4
- Win-rate impact on competitive deals: 31% to 82%
- Time-to-insight metrics: 8% to 71%
- Estimated revenue impact attribution: 9% to 53%
- Stakeholder satisfaction surveys: 58% to 84%
Output metrics do not disappear at stage 4. They move. They stay useful for running the team and stop being the story told to executives.
There is a related finding worth sitting with. Segment8 reports that the programs with the highest executive engagement share one trait: they have made themselves indispensable to a single recurring decision, such as the quarterly pricing review, the product council, the M&A pipeline, or the annual operating plan. Generalist programs without a flagship decision to support, they found, consistently struggle to justify their budgets.
That is a harder thing to fix than a metric. It is also the thing the metric is measuring.
One note on the numbers. The impact data here is self-reported by practitioners, and Segment8 publishes a competitive intelligence platform. The direction is consistent enough to act on, but read the percentages as indicative rather than precise.
What is a decisions-influenced register?
It is a spreadsheet.
Each row records one decision your work materially shaped: what the decision was, what the analysis contributed, which executive sponsored it, and the outcome where it is known. It is reviewed quarterly with that sponsor. Segment8 identified this as the single practice most associated with strong executive engagement in their research, and described its effect as turning the conversation about value "from anecdote into evidence."
One organization in their interview set keeps a register with 47 entries over 18 months, ranging from a $40 million pricing decision where the team supplied the competitor benchmark, to an acquisition abandoned after the team identified post-merger integration risk. Their CFO reviews it quarterly. That example comes from a single interview rather than from the survey, so read it as an illustration and not as evidence of a pattern.
Three things make the register unusually practical compared with the other measures on the list.
It requires no tooling
A shared sheet with five columns is the entire implementation.
It works backwards
Win-rate attribution and time-to-insight need instrumentation in place before the period you want to report on. A register can be backfilled. You can sit down this week and reconstruct the last two quarters from memory and your own calendar.
It survives challenge
A named decision with a named sponsor is checkable, which is exactly why it carries weight in a budget conversation. A number with no decision attached to it is not.
Two cautions. Entries have to be written while the decision is fresh, because six weeks later nobody remembers which analysis moved which meeting. And resist claiming sole credit for the financial outcome. The register is strongest when it describes what the analysis contributed to a decision other people made, which is both more accurate and harder to argue with.
Why do most teams never measure their value?
Not because it cannot be done.
Barnea's assessment is blunt: "Most firms which run active CI programs don't routinely measure their value." Some have built limited measures and then declined to use them systematically. His explanation is organizational rather than technical. It is not required by senior management, and it is not embedded in the culture.
He draws the consequence out one step further, and this is the part worth quoting to anyone who thinks measurement is bureaucracy. The lack of such measurement, he writes, "impacts the level of management support for the growth, development and influence of CI functions." The absence does not leave you neutral. It leaves you unfunded.
This is not unique to competitive intelligence. BCG's benchmarking of the insights function found that more than half of companies in industries such as consumer services, consumer packaged goods, and health and wellness did not regularly measure the return on their customer insight and commercial analytics spending. Different discipline, same gap, and that work dates from 2018, which tells you how long the problem has been in the open.
Where does a platform fit into this?
Mostly at the edges, and it is worth being precise about which edges.
Nothing on the register requires software. The entries are yours, the sponsor is yours, and the discipline is yours.
Where tooling does matter is in two of the other measures. Time-to-insight is hard to report if you cannot see how long a question actually took. And provenance has become a measurement issue in its own right: Segment8 found that as AI-generated competitor profiles proliferate, executives are asking for citations and confidence scoring, and that programs which cannot show their work are losing credibility. That puts weight on two things most buyers do not ask about: what a platform is contractually allowed to read, and whether it reads a whole document or a fragment of one. SinglePoint™ reads whole documents from licensed and internal content and links every claim back to its source, which is a different benefit from speed and increasingly the one under scrutiny.
If a platform is already in your 2027 plan, the evaluation is its own exercise. We have written buyer's guides for technology and for pharma and biotech, both of which set out eight criteria with three treated as pass or fail, and a scorecard to take into vendor demos. For a view of the field, see our comparison of the 12 best competitive and market intelligence platforms in 2026, and how SinglePoint compares against the alternatives.
For the argument from the other direction, built on customer outcomes rather than practitioner survey data, see how strategy teams are measuring market intelligence ROI.
What to do this week
Start the register. Five columns, last two quarters, reconstructed from memory. Then book twenty minutes with whoever signs your budget and walk them through it.
If the list is short, that is information too, and better to learn it in October than in a budget meeting.
Frequently asked questions
How do you measure the ROI of a market and competitive intelligence function?
There is no native conversion event for intelligence work, so direct ROI is rarely calculable. The practical approach is to measure decisions rather than outputs: maintain a register of decisions your analysis shaped, track win-rate impact on competitive deals where the data exists, and estimate revenue effect only where a specific decision supports the estimate. Programs at higher maturity consistently report these measures over counts of reports, briefs or alerts produced.
What is a decisions-influenced register?
A structured log, usually a spreadsheet, recording decisions where intelligence work materially shaped the outcome. Each entry captures the decision, the analytical contribution, the executive sponsor, and the eventual outcome where it is known. It is reviewed quarterly with leadership. Segment8's 2026 research identified it as the practice most associated with strong executive engagement.
Why is counting outputs a poor measure for an intelligence team?
Because it reports effort rather than effect. A count of reports produced tells a finance leader how busy the team was, not what would have happened without it. In a selective cost-reduction exercise, a high output count can read as volume available for trimming rather than as evidence of contribution.
What are CFOs planning to cut in 2026?
Gartner's 2026 Budget Assumptions survey of 142 CFOs and senior finance leaders found 64% planning for SG&A costs to grow more slowly than revenue. The functions most cited for reduction were human resources, corporate IT, legal and compliance, corporate finance, and marketing. Gartner characterized the approach as incremental cost containment that preserves essential capabilities, rather than across-the-board cuts.
Does measuring intelligence value actually protect a budget?
It makes the conversation easier rather than guaranteeing the outcome. Teams able to describe their contribution in terms of specific decisions tend to fare better in budget reviews than teams who can only describe their output. Much of the published evidence on this is self-reported, so treat it as directional.
Sources
- Gartner, 2026 Budget Assumptions survey, 142 CFOs and senior finance leaders, fielded August to September 2025, published 15 October 2025. Gartner newsroom
- Segment8, The State of Competitive Intelligence 2026, 612 practitioners across 41 countries, fielded July 2025 to March 2026. Segment8 research
- Avner Barnea, "Measuring the Value of Intelligence in Business: A Case Study," Competitive Intelligence Magazine Vol. 27, January 2022, hosted by SCIP. Read the paper
- BCG Henderson Institute, Measuring the ROI of Customer Insight. BCG Henderson Institute
