Real-world application of competitive intelligence

On 2 November 2016, the morning Microsoft announced Teams, Slack ran a full-page open letter to Microsoft in the New York Times. It welcomed the competition, agreed the category was real, and warned that a communication tool cannot be assembled by building Slack's feature list checkbox by checkbox. Slack could write that in a day because its product marketing team already knew what was coming and roughly how it would be priced. This lesson stays with that one rivalry, from the letter to the European Commission's investigation eight years later: what Slack read correctly, where the read failed, and what a company does when a competitive problem stops being solvable with messaging.

What Slack knew before Teams shipped

Microsoft's pattern was public. It had bought Skype in 2011 and Yammer in 2012, folded both into Office, and priced them as suite features rather than standalone products. It had also, by several credible reports, weighed buying Slack outright in 2016 for something in the region of $8 billion before deciding to build instead. Hiring posts, conference session titles and partner briefings all pointed the same way. Run the four-corners read the frameworks lesson walks through and the conclusion is not subtle: Teams would ship inside Office 365 at no incremental price. In March 2017 it did.

The useful part of that read was the pricing conclusion, not the launch date. Knowing a competitor is coming tells you little. Knowing it will arrive at a marginal cost of zero to your buyer tells you which of your existing arguments are about to stop working.

Sub-concept 1: what a bundled rival does to your battlecard

Slack's paid tiers sat at roughly $7 to $13 per user per month. Teams cost nothing extra for organisations already on Office 365, and in July 2018 Microsoft added a free tier, which was exactly where Slack's land-and-expand motion started. The persistent one-page profile the foundations lesson describes still mattered, but its content had to be rebuilt: the objection reps heard was now about procurement, not product. "We already pay for it" is not answered by a feature column.

What did work in Slack deals was the switching argument (more than 2,000 apps in its directory by 2020, so the integration surface, not the chat window, was the thing being replaced) plus the eventual footnote on "included": Microsoft raised commercial Microsoft 365 prices in March 2022, its first broad increase in about a decade. Bundled is not the same as fixed, and a card that never says so hands the decision to the CFO unchallenged.

Sub-concept 2: two user counts that were never comparable

Microsoft disclosed Teams daily active users: 13 million in July 2019, 20 million that November, 75 million by April 2020, 145 million a year later. Slack's last comparable public figure was around 12 million daily users in early 2020. Side by side the story wrote itself, which is the point. A Teams seat could be switched on by one administrator for 40,000 employees. A Slack seat had to be chosen and paid for.

Slack's counter-metric was commercial: roughly 88,000 paid customers at its June 2019 direct listing, about 575 of them worth more than $100,000 a year, and revenue just over $900 million in the year to January 2021, growing above 40%. Both sets of numbers were true at once, and the press ran the daily-user comparison anyway. The second-order cost is the one to plan for: once a rival has picked the scoreboard, you answer for their metric in every board meeting, analyst call and enterprise deal review. Publish yours early or inherit theirs.

Sub-concept 3: the gap a bundle cannot cover

An Office licence covers your employees. It does not cover your agency, your outsourced support vendor or the finance team at the company acquiring you. Slack pushed at that seam with Slack Connect in June 2020, promoting shared channels from a feature to the headline: work across company boundaries, where the bundle has no claim.

Where the read failed was video. Slack's calls capped at 15 participants, and by March 2020 meetings had become the way collaboration tools got bought. Slack answered with a partnership rather than a build, signing an AWS agreement that June which put Amazon Chime behind its calls while Amazon deployed Slack internally. A positioningpositioningThe mental space you want your brand to occupy in your target customer's mind relative to alternatives.View full definition → read can be right and still lose the quarter, because a feature you filed as a checkbox can become a procurement requirement in about six weeks.

Sub-concept 4: when the answer stops being marketing

On 22 July 2020 Slack filed a complaint with the European Commission, arguing that Microsoft was illegally tying Teams to Office, installing it by default, making removal difficult and obscuring its real cost. That is what escalation looks like when the levers marketing controls have all been pulled and the buyer's default still is not yours.

Note what the filing required: dated evidence of install behaviour, packaging changes and pricing across years. That is a competitive archive, and a team that only ever keeps the current picture cannot produce one.

The timeline is the harder lesson. Salesforce agreed to buy Slack five months later for $27.7 billion, closing in July 2021. Microsoft began selling Office without Teams in the EEA and Switzerland in October 2023 and globally in April 2024. The Commission opened formal proceedings in July 2024, four years after the complaint, and Microsoft subsequently offered commitments on pricing and interoperability. By then Slack had been inside Salesforce for three years and Stewart Butterfield had left. Regulatory remedies do not run on campaign time; if your plan depends on one, you need a plan for the years in between.

Real-world case 1: the two-day response

The 2016 letter cost a full page of newspaper inventory and slowed Teams adoption by nothing measurable. Judge it on what it could actually move: it set the frame before Microsoft could, and it gave every rep and every journalist covering the launch one sentence to reachreachThe number of unique people exposed to your message in a given period. Unlike impressions, reach counts each person once, no matter how often they see it.View full definition → for. Microsoft, for its part, was running the same discipline in reverse; its internal list of software prohibited for employees included Slack's free and paid tiers, as reported in 2018. Fast responses buy framing, not share. Budget them accordingly and stop expecting a pipelinepipelineAll active sales opportunities across the stages of the sales process, together with their combined potential value and probability of closing.View full definition → number from them.

Real-world case 2: the read was half right

Slack kept winning large deployments through the Teams era, including IBM at more than 350,000 seats, and kept growing revenue at over 40% a year. Teams passed 250 million monthly active users by early 2022. Both facts hold. Bundling did not stop Slack's revenue line; it capped how much of the category Slack could own, and the endgame was a $27.7 billion sale into a larger suite rather than a standalone winner. When you model a bundled rival, model that outcome explicitly: growth continuing while strategic position erodes is the most common shape, and the one most CI decks fail to draw.

CMO action items

• Keep a dated archive, not just a current view: screenshots of competitor pricing pages, packaging tiers, default install behaviour and terms, filed monthly. Slack's 2020 complaint rested on a multi-year record that no quarterly refresh would have produced.
• Re-cut sales material around procurement the moment a rival bundles: drop the feature columns, answer "we already pay for it" with switching cost and the price history of the suite, in buyer language taken from the interview discipline the foundations lesson sets out.
• Choose and publish your own scoreboard before the rival publishes theirs, and brief finance and IR on it, since that is where the comparison gets made.
• Write the escalation ladder in advance: which competitive moves get a same-week message, which get a roadmap change, which go to legal or regulatory review, and who signs off on each.

Common mistakes that kill results

• Answering a bundle with a feature comparison. The buyer's question is why they should pay twice, and a comparison grid never answers it.
• Reading a rival's activation numbers as demand. Administrator-enabled seats and paid, chosen seats are different quantities, and treating them as one leads teams to concede a fight they are still winning commercially.
• Dismissing a feature as a checkbox because your positioning says it is one. Slack's 15-participant calls looked like a non-issue until meetings became the buying trigger.
• Treating a regulatory complaint as a competitive strategy. It is a slow instrument with an uncertain outcome, useful alongside a commercial plan and dangerous as a substitute for one.