Hypothetical example: A company is entering a new market. The campaign is live, the sales team is holding meetings, and the local partner is getting ready. At the first review, sales are coming in more slowly than expected. Marketing asks for more visibility, sales asks for more time, and operations asks for product adaptations. Each proposal has a reasonable rationale of its own.
The meeting ends with a decision to “give it a little longer.” At the next meeting, the definition of success changes: the initial expectation of sales gives way to discussions of meeting volume, followed by brand awareness. Because the conditions for stopping the investment were never written down, every review produces a new reason to continue.
The questions facing management now expand: Is there an opportunity in this market? How much more budget and team capacity will we commit to exploring it? Answering the first question positively does not justify leaving the second open-ended.
At Sellf, the model we propose for growth engineering makes this distinction visible from the outset: the Walk-Away Line. Before spending begins, every initiative specifies the evidence that will trigger stopping, the evidence that will trigger scaling, and when the decision will take effect. This establishes boundaries for both learning and resource allocation.
1. The Rationale for Continuing Quietly Changes
“We’ve invested too much to stop now” turns past spending into a justification for future spending. Yet the decision management faces today concerns the expected return on the resources committed from this point forward.
Sunk costs are only part of the issue. Across four experiments in a study published in 2018, Feldman and Wong found that framing continuation as “action” and withdrawal as “inaction” increased the tendency to maintain commitment to a failing investment.
The management implication we draw is this: writing down what will happen after an initiative is stopped may make the decision easier to implement. When the budget and capacity released have a defined next destination, management can see a concrete alternative.
In the new-market example, “everything will have been wasted if we exit” and “we will redirect the localisation budget to a bottleneck in our existing market” create different grounds for a decision. The economic value of the Walk-Away Line also lies in the resources it releases.
2. Define the Decision in Four Lines Before Spending Begins
In the model we propose, four lines are added at the start of every growth initiative. The explanations can be detailed; the core decision stays brief.
1. Hypothesis: What Are We Assuming to Be True?
“We will grow in the new market” expresses an intention. “With local delivery and sales support, our existing product can acquire customers profitably in this market” is a testable assumption. The hypothesis explains the commercial mechanism through which the investment will create value. The same logic applies to team investment: specify which capacity constraint the new role will remove and how that will translate into an economic outcome.
2. Metric: Which Single Number Will Judge the Initiative?
The primary decision metric is tied to the initiative’s economics: incremental contribution profit, contribution margin, or the payback period on customer acquisition investment, for example. ROAS and impressions cannot fulfil this role. If incremental contribution profit is chosen for the hypothetical market entry, the treatment of product, logistics, returns, marketing, and additional market-specific operating costs is defined in advance. Sales diverted from existing markets are also separated out. Supporting measurements can be tracked alongside the single decision metric to help diagnose problems.
3. Thresholds and Date: What Results Will Trigger Stopping or Scaling?
Stopping and scaling thresholds are written down for the metric, together with an explicit calendar date. A low contribution profit may call for stopping; with payback period, the direction is reversed. Thresholds are derived from cash capacity, costs, and the expected return on the alternative investment. The “monitor” zone between the two thresholds also has an end date and a spending cap. When that period ends, spending stops unless renewed approval has been granted. This prevents uncertainty from turning into indefinite additional funding.
4. Owner and Next Move: Who Implements the Decision, and Where Do the Resources Go?
A single owner with decision-making authority is designated. If the initiative stops, the spending to be closed, the work to be transferred, and the destination of the released capacity are specified. If it scales, the approved budget increase and operational capacity are also defined. Repeatedly sending growth back for approval after the evidence arrives can delay the return on the opportunity. The scaling threshold makes management’s responsibility to act on positive evidence visible as well.
3. The Line Can Only Work Once the Evidence Has Matured
Putting a date in the calendar does not, on its own, guarantee that a sound decision can be made on that date. If the sales cycle is incomplete or returns have not yet become visible, it may be too early to assess profitability.
Experiment checklists developed by Microsoft researchers emphasise defining duration and sample size in advance; they explain how ending a test as soon as the results look favourable can produce an incorrect decision. This finding concerns online controlled experiments. The principle we draw for broader growth initiatives is to design the decision date around the time needed for evidence to develop.
The third line therefore also includes the conditions for sufficient observations, the measurement source, and how uncertainty will be assessed. Any limited extension for insufficient data is defined in advance; without that authority, no additional budget is released. Separate stopping limits that protect customer experience or cash are also monitored. Microsoft’s experimentation approach likewise distinguishes the primary evaluation criterion from guardrail metrics.
The Marginal Unit Equation evaluates an existing unit; the Walk-Away Line precommits the resource decision for an initiative that has not yet begun.
4. Add Four Lines to the Next Budget Meeting
Our single practical recommendation is this: at the next budget meeting, attach this four-line form to every request for a new campaign, channel, market, launch, or team investment. Then check it against three questions:
- Was the line written before spending began?
- Does the line have a date and an owner?
- Does crossing the line become another meeting agenda item, or does it trigger a decision that has already been approved?
Here, acting automatically means that the owner implements the decision once the agreed conditions are met. If a threshold is revised later, the new evidence and rationale are recorded; the old result is not declared a success against a new target.
Can you specify today the result and date that would trigger a stop to your next growth initiative?
If you’re wondering where your growth model stands, we can discuss it in 30 minutes.
