CMO playbook & advanced tactics for product launches

Your product organisation will ship more next year than the market can absorb. Salesforce puts out three platform releases a year, each carrying hundreds of changes, and almost none of them get a launch. That gap is the CMO's problem, not the product team's. Attention is the one input you cannot buy more of at the margin: the press cycle, the analyst calendar, the field's capacity to learn a new pitch, the customer's willingness to open one more email. All of it is fixed for the quarter. Everything else in a launch plan scales with budget. So the leadership job here is arbitration, and the price of getting it wrong is paid by the launch you underfunded, not the one you overfunded.

Core concept: one attention budget, many launches

Run forty launches a year and treat each as important and you have run none. The portfolio question is not "is this product good", it is "what does this product displace". Every top-tier launch you approve pushes something else down the calendar, and the cost of that displacement never appears in the launch's own business case.

Four inputs decide the arbitration:

• Twelve-month ceiling, not lifetime valuelifetime valueLifetime Value: the total revenue (or profit) a customer generates throughout their entire relationship with your business.View full definition →. Lifetime value flatters everything. Ask what this can book by the same quarter next year.
• Perishability. A competitor's roadmap, a hardware cycle, a retail season or a regulatory date can close the window. Non-perishable launches wait; that is what makes them cheap to defer.
• Enablement cost. Hours of sales, support, partner and legal retraining. This is close to fixed per launch and largely independent of the revenue at stake.
• Reversibility. You can re-run a campaign. You cannot re-launch. The market writes the story once.

The uncomfortable arithmetic: a top-tier launch consumes roughly the same executive attention whether the product is worth five million or two hundred million. Only expected value varies, so only expected value should decide.

Sub-concept 1: arbitrating between tiers

Assuming the tier definitions the foundations lesson sets out, the leader's decision is not which tier a product belongs to. It is how many top-tier slots exist at all, and how far apart they sit. Run two of them at the same audience inside six weeks and the second inherits the first's leftovers: the same journalists, the same analyst briefings, the same field reps who have not finished learning pitch one.

Tier inflation is the standard failure mode. Every general manager argues their product is top tier, and if the tier is defined by a threshold rather than a quota, the threshold moves until the label means nothing. Cap it as a quota (a fixed number of slots per quarter per audience), publish the calendar two or three quarters out, and make the trade explicit: taking a slot means someone else loses one, by name, in the room.

Then hold back a reserve. Ten to fifteen percent of the quarterly launch budget, unallocated. The arbitration error that hurts most is not overspending on a small launch, it is the launch the market decides is big while you resourced it small. When that happens you have days to react, and no reserve means the answer is no.

Sub-concept 2: launching when you cannot supply

Sony shipped the PlayStation 5 on 12 November 2020 into a semiconductor shortage, sold out immediately, and stayed sold out for the better part of two years. About ten million units moved in the first eight months, the fastest start of any PlayStation, with demand generationdemand generationMarketing activities designed to attract and capture contact information from prospects interested in your offer, creating a pipeline of potential customers.View full definition → contributing nothing to that number.

When supply is the binding constraint, marketing spend shifts from acquisition to allocation. The work becomes registration lists, direct-to-consumer invite queues, restock communication cadence, retailer bundling rules and scalper suppression. Sony ran invitation-based sales through PlayStation Direct precisely because an open queue rewards bots. Every impression you buy above your supply line converts into a resale listing and a customer who blames you rather than the shortage.

The second-order cost is durable. Scarcity that lasts one quarter builds anticipation. Scarcity that lasts eight quarters trains buyers to stop checking, hands attach-rate revenue (accessories, software, subscriptions) to nobody, and lets a competitor own the shelf by simply being available.

Sub-concept 3: the cost of concurrency

IKEA introduces around 2,000 new products a year. Almost none of them are launches. The annual range change concentrates the news into one window, and the rest arrive as a price label on a shelf. That concentration is a choice about internal capacity as much as customer attention.

Launches bill the same fixed departments every time: legal review, support macros and training, partner enablement, translation, packaging, retail merchandising. Those teams do not scale with the marketing budget. Overload them and the damage shows up as launch debt: the follow-through on launch three (the case studies, the win/loss reads, the pricing corrections at week six) gets cut to staff launch four. Run a year like that and your top-tier launches quietly produce mid-tier outcomes, with no single decision you can point to as the cause.

[VIDEO: "How Apple Launches Products" - youtube.com/watch?v=lBG11tajL_4 - Why watch this now: it shows a company compressing a year of product news into a small number of dated moments, which is the concentration decision this section asks you to make.]

Sub-concept 4: deciding what not to launch

Most of the arbitration is demotion, and demotion reads inside the company as a verdict on the team. Defuse that by tying tier to the published expected-revenue rule rather than to advocacy, and by running the readiness check the frameworks lesson sets out before the tier is confirmed, not after the date is booked. A product that fails readiness and keeps its date does not get a smaller launch, it gets a bad one.

Write a kill rule too. If a launch has slipped twice and its twelve-month ceiling has not risen, it ships as a release note and the slot goes to the next candidate. Without that rule, the calendar fills with things nobody will defend and nobody will cancel.

Real-world cases

Case 1: Salesforce, one story a year

Salesforce sells CRMCRMCustomer Relationship Management: software and strategy to manage and analyse customer interactions throughout their lifecycle.View full definition → and marketing software, so its own launches double as product demonstrations, which raises the stakes on the arbitration. The structure is deliberately narrow: three seasonal platform releases carrying the bulk of the shipped work, and one September moment at Dreamforce that carries the company narrative. Einstein got that slot in 2016; Agentforce got it in 2024. Everything else, including features that would headline a launch at a smaller company, rides the release notes. The cost is real: individual teams watch their work land as a bullet point. The return is that when Salesforce says something is important, buyers and analysts still believe it.

Case 2: sony and the PS3, scarcity without demand

The PlayStation 3 arrived in November 2006 at 499 and 599 dollars in the US, constrained by blue-laser diode yields, with the European launch pushed to March 2007. Sony sold each unit below cost for years. The supply shortage looked like the PS5 situation and behaved nothing like it, because demand at that price was thin and the launch software lineup was thinner. Nintendo and Microsoft took the generation's early lead. Scarcity only creates urgency when buyers already want the thing at the price; otherwise a constrained launch is just a slow one, and every week of constraint is a week your competitor spends unopposed.

Case 3: ikea and the end of the catalogue

The IKEA catalogue ran for 70 years and peaked at roughly 200 million copies. It was also a rationing device. The cover, the room sets and the page count were finite, so somebody decided each year which of the 2,000 new products got a spread and which got a line. IKEA retired it in December 2020. Digital channels have no page count, which removes the forcing function along with the printing bill. Any leader moving an always-on portfolio off a physical container has to reimpose an artificial ceiling, because a channel with no scarcity will not arbitrate for you.

[VIDEO: "Building a Go-To-Market StrategyGo-To-Market StrategyThe strategy defining how you'll launch a product: target segments, channels, value proposition and coordinated action plan.View full definition →" - youtube.com/watch?v=3Fdzfrx0t74 - Why watch this now: it itemises the GTM components every launch consumes, which is what you need in order to price a top-tier slot against a lower one.]

CMO action items

• Set the number of top-tier slots per quarter per audience before you see the candidate list, and publish it. A quota survives lobbying; a threshold does not.
• Keep ten to fifteen percent of the quarterly launch budget unallocated so you can promote a launch the market has already decided is bigger than you thought.
• Cost every launch in enablement hours as well as media dollars, and get the support and field leaders to sign the number. That is the figure that caps your calendar.
• For any supply-constrained launch, agree the allocation mechanism (queue, invite, waitlist, retailer rules) before the date, and cut acquisition spend the day demand passes supply.

Common mistakes that kill results

• Launching everything at once. Three top-tier launches in a quarter is one top-tier launch and two that were mishandled in public. The failure is invisible in the plan and obvious in the results.
• Buying demand you cannot fulfil. Above the supply line, marketing spend funds resellers and manufactures resentment. Reallocate it to allocation fairness and restock communication.
• Treating tier as status. Once demotion is read as an insult, tiering becomes a negotiation with the loudest GMGMGross margin is the share of revenue left after subtracting the direct cost of producing goods or services, expressed as a percentage of revenue.View full definition → and stops being a resource decision.
• Skipping the post-launch tail. Weeks four to twelve are where pricing gets corrected, objections get answered and case studies get built. Those weeks get cut first when the next launch is already late, and the compounding stops there.