How to Run a Platform-Change Drill
A fire drill happens before the building burns; a platform drill, before the gate narrows. The drill’s question is simple: if our most critical rented gate narrows seriously within six months — the feature vanishes, the algorithm turns, the account gets suspended — what do we do in the first week, and what do we keep alive through the first month? Working that question for two hours a year, at a calm table, turns the crisis day into the plan’s day rather than panic’s.
Below is the drill’s construction: choosing the scenarios, running the two-hour table, and writing the output — a one-page response card. The aim is not prophecy; it is the preparation itself. Which scenario arrives is unknowable, but the prepared business exercises the same three muscles in all of them: detection, substitution, communication.
What Is the Problem?
BU BÖLÜMÜN ÖZETİ
- It notices late
- It decides in anger
- It holds no substitute
The undrilled business loses the narrowing day three times over.
It notices late
A narrowing usually arrives not as an announcement but as a quiet slope: traffic seeps away for weeks, and with no alarm threshold set, someone calls it seasonal. By the time it is noticed the loss has accumulated — and the diagnosis argument crowds out the treatment. The first loss is not time; it is the failure to notice time passing.
It decides in anger
The unprepared response runs a script: first bargaining, then anger — the industry’s public blocking and boycott debates are the visible version, as in the publisher standoffs reported at gHacks. Anger is understandable and is not a strategy — and the spectacle of giants living the crisis in public shows why even a small business should live it at a table instead. The first decision made on a crisis day is usually the decision later reversed.
It holds no substitute
The channel meant to replace the narrowed gate cannot be built that day: a list does not gather in a day, a second gate does not mature in a week. That is the drill’s real function — hauling the substitution need from the crisis day back to today, because a registered relationship accrues only with time.
Why Does It Happen?
BU BÖLÜMÜN ÖZETİ
- Disaster is filed under unlikely
- Preparation is filed under pessimism
- Nobody owns it
The missing drill culture has three causes.
Disaster is filed under unlikely
“A platform that size wouldn’t do that” gets said against the evidence — yet feature dismantlings are procedure, not exception, and arrive without reasons. The question is not “would they” but “what does it cost us when they do”; the drill converts a probability argument into a cost calculation.
Preparation is filed under pessimism
Working a crisis scenario feels contrary to growth culture and stays off the table. The truth is inverted: the boldest growth is the luxury of the business with a downturn plan — working brakes earn the right to accelerate, in channel strategy as elsewhere. The drill is not pessimism; it is courage’s infrastructure.
Nobody owns it
Platform risk is everyone’s topic and no one’s job: marketing watches traffic, sales watches revenue, management watches totals — nobody watches the gate itself. The drill demands one name: the person who updates scenarios, watches alarms, and convenes the table once a year.
How Is It Done?
BU BÖLÜMÜN ÖZETİ
- Step 1: draw the scenario from the inventory
- Step 2: run the two-hour table
- Step 3: write the response card, date it
The drill is three steps; two hours total, one page of output.
Step 1: draw the scenario from the inventory
The dependency inventory’s reddest cell is taken and written at three severities: mild (a feature narrows — visibility drops), medium (an algorithm or rule change — the gate halves), severe (account or access loss — the gate closes). The scenario is written observably: in which gauge, at what number, does it show? A rehearsal of observation, not prophecy — the scenario’s value lies not in its probability but in the muscle it exercises.
Step 2: run the two-hour table
Three questions, in order. Detection: in how many days, by which gauge, would we catch this narrowing — what is our alarm threshold? Substitution: which channel takes the first week’s weight, what do we tell the registered list, which spend stops? Communication: to customers, to the team, to partners — who says what, when? Each answer is compressed to one sentence; if the debate runs long, what has been found is not an answer but an uncertainty — and that uncertainty is the drill’s true finding, written straight onto the homework list.
Step 3: write the response card, date it
The output is one page: scenario, alarm threshold and gauge, first-week steps, first-month priorities, communication lines, and the owner’s name. The card lives beside the inventory and is reworked once a year, when the inventory refreshes. An undated card is an undrilled plan: the platform changes, the business changes — the card must not age with them.
How Long, at What Cost?
BU BÖLÜMÜN ÖZETİ
- Two hours of table a year, five minutes of alarm a quarter
- The dearest output is the cheapest discovery
- By-product: daily decisions sharpen
The drill’s economics are insurance economics.
Two hours of table a year, five minutes of alarm a quarter
The build is one two-hour meeting; the operation is a five-minute quarterly reading of the alarm gauge. Against that cost stands the unprepared crisis week — panic decisions, unstoppable spending, delayed communication. The ratio is not open to debate.
The dearest output is the cheapest discovery
Most tables discover the same things: the substitute channel is weaker than assumed, the registered list smaller, the contact chain missing entirely. Made on the crisis day, that discovery is invoiced; made at the drill, it becomes homework. The drill’s profit line is every discovery hauled from the crisis day back to today.
By-product: daily decisions sharpen
A team that has walked the scenario once reads everyday choices differently: “which gate does this investment deepen our dependence on” starts asking itself. The drill is less an annual meeting than a reflex that settles into the decision language.
The Common Mistake
BU BÖLÜMÜN ÖZETİ
- Softening the scenario
- Writing the card, skipping the alarm
- Mistaking the drill for an exit plan
The drill runs; three leaks drain it.
Softening the scenario
Lowering the severity at the table — “it wouldn’t get that bad” — turns the drill into a wish. The severe scenario is worked even if it never arrives; fire drills are held by people who do not expect a fire. The muscle trains independent of the probability.
Writing the card, skipping the alarm
A threshold written but on nobody’s calendar is an alarm that never rings. The alarm binds to a name and a date; the canary gauges join the same calendar — sector data sings before your own does.
Mistaking the drill for an exit plan
The drill is not a “let’s leave the platform” decision; the rented gate stays valuable, narrowing risk included, and stays in use. The aim is not to exit the gate but to stop being hostage to it — the many-door arrangement is the order of living with platforms, not without them.
Frequently Asked Questions
Sık Sorulan Sorular
One table a year suffices, plus an interim round whenever the inventory’s red cell changes. More frequent drills breed fatigue; rarer ones get forgotten — the annual rhythm balances the card’s freshness against the team’s patience.
With yourself and a sheet of paper: the three questions, answered in writing, in the same order; one hour instead of two. The solo business’s advantage is speed, its risk is blindness — a written answer shows your own assumptions back to you. Once a year, if possible, a mutual read with a trusted peer.
The order is fixed: first the alarm threshold — a business that cannot see cannot do anything; then the registration arrangement — a substitute is born only by accrual; new gate experiments come last. Fragility does not close in a day; but its direction turns in a day, and that turn is precisely the drill’s job.