CMO playbook & advanced tactics for competitive intelligence

The first competitive intelligence hire is usually not a hire at all. A CMO adds "competitive" to a product marketer's job description, gives it maybe a fifth of their week, and waits for win rates to move. Two quarters on, the cards are stale, no buyer has been interviewed, and nobody can say whether last quarter's repositioning changed a single deal. Nobody decides whether to do CI. What a CMO decides is how many heads, with what authority, and inside which legal limits.

What the function costs, and what a wrong read costs

Three staffing models, in ascending order. Embedded: 20 to 30 percent of one product marketer, no new tooling. This holds up to roughly three rivals that matter and a sales team you can brief in one room. Dedicated: one full time CI lead in product marketing plus a subscription stack that runs into the tens of thousands of dollars a year. Team: a lead, an analyst who owns data hygiene, and an outsourced interviewer for buyer debriefs. Crayon, which sells CI software and therefore has a view on the answer, has surveyed the discipline annually for years, and the dedicated headcount it reports stays in the low single digits per company even at large firms. Any plan that opens with five people will not get funded.

Run the arithmetic once before you ask. Four hundred competitive deals a year at a 120,000 dollar average contract value: two points of win rate is just under a million dollars of new ARRARRAnnual Recurring Revenue (ARR) is the normalized, predictable revenue a subscription business expects to earn from active contracts over a single year.View full definition →. That is the number the headcount case gets measured against, which is also why the function belongs to whoever owns pipelinepipelineAll active sales opportunities across the stages of the sales process, together with their combined potential value and probability of closing.View full definition → rather than to a research group filing reports.

The second thing a CMO owns here is the limit. Pulling a rival's pricing off a public page is research. Having someone pose as a buyer to sit through their demo is a grey zone most legal teams will not sign off, and impersonating a real person to extract records is a crime in several jurisdictions: Hewlett-Packard's 2006 pretexting affair, where investigators impersonated journalists and directors to obtain phone records, cost the chairman her seat and produced criminal charges. Trade secret theft has its own statutes, the US Defend Trade Secrets Act of 2016 and the EU Trade Secrets Directive among them. Write a one page policy, have legal approve it, make everyone in marketing and sales sign it. The cheapest control in it is a single rule: no one on your team ever misrepresents who they work for.

Sub-concept 1: the battle card as a revenue asset

Assume the one page sales asset the foundations lesson defines, and start from the fact that most of them are never opened. The usual failure is format. A feature comparison table is useless in a live call because buyers are not adjudicating parity, they are managing the risk of choosing wrong and having to explain it internally.

HubSpot's positioningpositioningThe mental space you want your brand to occupy in your target customer's mind relative to alternatives.View full definition → against Salesforce in the SMB segment is a useful model partly because so much of it is public: HubSpot runs comparison pages on its own site rather than burying the comparison in a sales-only PDF. The cards worth building answer four things. What does the buyer fear losing? Where does the rival genuinely win (say it plainly, because reps who never hear a concession stop trusting the card)? What is the one reframe that changes the comparison? And what proof point survives a procurement review?

Two governance rules. Card only competitors appearing in more than about five percent of qualified pipeline; a library of forty cards means reps use none of them. And write every card assuming a prospect will screenshot it, because eventually one will. Anything you could not defend in a comparative advertising complaint, and EU rules require comparative claims to be verifiable and not misleading, does not belong on the page.

Sub-concept 2: share of voiceshare of voiceYour brand's share of total advertising or conversation volume in your category, measured against competitors over a defined period.View full definition → tracking as a leading indicator

The claim that share of voice leads market sharemarket shareThe percentage of total industry sales your company captures in a given period. It measures competitive position relative to rivals in a defined market.View full definition → has a specific source: the IPA databank work by Les Binet and Peter Field puts it at roughly half a point of share growth a year for every ten points by which share of voice exceeds share of marketshare of marketServiceable Obtainable Market: the share of your SAM you can realistically capture given current resources, channels, and competitive position.View full definition →. Ten points is expensive to fund, which is the arbitration nobody enjoys. Buying excess share of voice against a better capitalised rival is a losing trade.

Samsung is the cautionary case at the top end. It has been among the world's largest advertisers for over a decade, reported to have spent north of ten billion dollars on marketing in 2013 alone, and it bought smartphone share with that spend. Very few CMOs can answer that with money, so they answer it by narrowing the category they measure themselves in.

Track with Brandwatch, Semrush or SparkToro. A rival's share of voice moving fifteen points in a quarter means something happened: funding, a launch, a campaign, a repositioning. The mistake is treating the spike as an instruction. Often the right call is to hold and watch it decay, because paid-driven share of voice with no product news behind it is gone within weeks.

Sub-concept 3: win/loss analysis as competitive truth

The buyer debriefs the foundations lesson describes give you the highest quality input available, and they are also the easiest thing in the programme to run badly. Four failure modes:

• The account owner runs the interview. Buyers will not tell the rep who lost the deal why they lost it. Use a neutral interviewer; outsourced programmes tend to price per interview in the high hundreds to low thousands of dollars, which sets a natural ceiling on volume.
• Survivorship bias. Won accounts answer, lost accounts ghost. Track the response rate on losses separately and publish it next to the findings, or the sample quietly tilts toward happy buyers.
• CRMCRMCustomer Relationship Management: software and strategy to manage and analyse customer interactions throughout their lifecycle.View full definition → loss reasons treated as evidence. Reps code "price" because it is the least uncomfortable box. Interviews routinely surface something else: implementation risk, a champion who left, a budget that never existed.
• No-decision deals filed as "lost to competitor X". Code them apart or you will build messaging against a rival who never won anything.

Twenty to forty structured interviews a quarter is enough to see patterns in most mid-market businesses. Surveys do not substitute: they capture what the buyer remembers, not the sequence that produced the decision.

Sub-concept 4: competitor messaging deconstruction

A rival's homepage, pricing page and ad library tell you what they believe their buyer cares about, and changes to them are dated evidence. Samsung's move from the direct attack posture of "The Next Big Thing Is Here" in 2012 and 2013, whose ads staged queues outside Apple stores, to the broader "Do What You Can't" line in 2017 showed up in the creative long before anyone briefed it as strategy. Read as a signal, it said Samsung had stopped fighting for switchers and started defending a brand.

Sources: the Wayback Machine for page history, the Meta Ad Library, Google's Ads Transparency Center, LinkedIn's ad transparency page. The EU Digital Services Act now obliges the largest platforms to maintain public ad repositories, which makes this cheaper than it used to be. Put one named person on a monthly review of the top three rivals, and require the output to be a diff, what changed since last month, not a fresh description of their website.

Real-world cases

Adobe and Figma, 2022 to 2023: Adobe agreed in September 2022 to acquire Figma for about 20 billion dollars and abandoned the deal in December 2023 after the UK Competition and Markets Authority and the European Commission signalled they would not clear it. Adobe paid a one billion dollar termination fee. The regulators' case turned on how closely the two products competed, which is the same question Adobe's own competitive material had been answering internally for years. Everything your team writes about a rival is discoverable and can be read back to you as evidence of market definition. That is a real constraint on how battlecards and launch memos are worded at scale.

HubSpot's free tier: HubSpot launched a free CRM in 2014 and has built its platform around it since, competing with Salesforce from below rather than on feature parity. The second-order effect on intelligence is one most teams miss. Once a free product exists, "lost to a competitor" and "did not buy" start to blur, and a prospect parked indefinitely on the free tier looks like neither. If your loss taxonomy has no code for that, your competitive reporting will overstate rival wins.

Samsung and Apple: the 2012 US jury verdict awarded Apple over a billion dollars, reduced on appeal, and the dispute ran until a settlement in 2018. Samsung's own internal comparison of its handset against the iPhone, a long feature by feature document, became an exhibit. Through all of it Samsung's component business kept supplying Apple with displaysdisplaysThe total number of times an ad or piece of content is displayed, regardless of clicks. Each display counts as one impression, even to the same person.View full definition → and memory. Two lessons for a CMO: your competitor can also be your supplier, which means intelligence needs firewalling by business unit, and comparative claims written by marketing can end up in front of a court.

CMO action items

• Name the owner and the budget line this week, and pick a model out loud: embedded, dedicated, or team. Ambiguity here is what produces stale cards two quarters later.
• Get the ethics and legal policy written, approved and signed, and make sure it answers the specific question of whether anyone may book a demo with a rival under a false employer.
• Instrument the loss side properly: neutral interviewer, separate code for no-decision, and the loss-interview response rate reported next to every finding.
• Run a weekly brief to sales leadership and a monthly deeper read for product and the exec team. Each ends with a recommended action and a kill list: competitors you are going to stop tracking.

Common mistakes that kill results

• Treating CI as a static deliverable. Gartner's research on B2B buying puts the share of the purchase process that buyers spend with all potential suppliers combined at roughly 17 percent, across a buying group that often runs to six or ten people. The comparison happens in documents you never see, which means your material has to travel without a rep attached and stay current between quarterly reviews.
• Building for internal comfort. If the output never changes homepage copy, a sales script or a campaign brief, it is reporting. Tie each monthly analysis to one artefact that shipped because of it.
• Ignoring the status quo. In most B2B markets the commonest lost deal is inaction: renew what exists, defer, or build internally. A "vs. do nothing" card, aimed at the cost of delay rather than at any rival, usually covers more pipeline than the branded cards do.