Phase plan template for Ecommerce

The phase plan with the limits a store needs: authority bounded by catalogue segment rather than by campaign, a margin floor no change may cross, stock cover checked before weight is added, and promotion weeks held outside ordinary permissions.

Everything else is the general phase plan, unchanged. One row per area per phase, an explicit approval list, a reversal rule and a record of every change.

Write it when somebody other than the owner starts making changes in a trading account, which in a store is usually within a month of the first agency or first hire. The pace of retail makes a fully approved workflow impractical and makes an unbounded one expensive.

Revisit it before every peak, because peak weeks deserve narrower limits rather than wider ones.

A trading account changes constantly: stock positions move, promotions open and close, delivery cut-offs arrive, competitors discount. A workflow where every change waits for approval will not keep up, and the account drifts while the approvals queue.

The usual response is to hand over the account entirely, which works until the first expensive week and then becomes a conversation nobody can resolve, because there is no record of what was changed or why.

A phase plan in a store is the practical middle. It gives the person closest to the account enough authority to keep pace within an ordinary week, bounds it in the terms that actually matter here, and widens it on evidence rather than on how long everyone has worked together.

The most useful limit in a store is not a number, it is a boundary.

Moving budget between campaigns inside one catalogue segment is a tactical change: same margin, same stock, same customers, and the person closest to the account is better placed to make it than anybody else. Moving budget between segments is a commercial decision, because it changes which part of the range the quarter is about, and it has consequences for stock, for buying and for the plan.

So phase one grants within-segment authority and requires approval for anything across segments. That single line removes most of the friction without granting anything that changes the shape of the business.

No permission should allow weight to be added to a segment whose margin is unknown.

This is stricter than it sounds and it is deliberate. A store's account always contains segments whose contribution nobody has calculated, and those are exactly the segments that look best on revenue-based reporting. Authority that permits money to flow towards them will reliably move it in the wrong direction, one reasonable decision at a time.

Writing the margin condition into the limit column also has a useful side effect: it creates pressure to fill the margin column in the media plan, because nobody can use their authority on a segment that has not been costed.

The one failure spending cannot recover from is advertising a product nobody can buy.

So the permission carries a condition rather than a limit: weight may be added only where cover has been checked that week and recorded. The check is a minute and it prevents the specific case where a campaign at full weight runs through a sell-out over a weekend.

For the same reason, decreases usually need no condition at all. Reducing weight on a segment is reversible and cheap, and a phase plan that requires approval to reduce is one that will be worked around in an emergency.

The instinct during a peak is to loosen authority, because things are moving quickly and approvals are slow.

That is the wrong direction. Peak weeks carry several times the ordinary spend, audiences saturate within days, and the usual signals are least reliable precisely when they are being read most often. A limit that is sensible in a normal week — a fifth of a campaign, once a week — is a much larger amount of money during a peak and a much faster way to exhaust an audience.

The workable arrangement is to plan the peak in advance, with the weights agreed in the flight plan, and to hold ordinary permissions outside the promotion weeks. Anything that has to change during them is a decision with a person attached, which during a peak means a phone call rather than a queue.

Automated rules belong in this document and they take the same columns, with one addition: the rule's behaviour is tested against the last two peaks before it is given live authority.

Rules calibrated on ordinary weeks behave badly in the weeks that matter. A rule that moves budget away from a rising cost per order will, during a peak, move money out of the best week of the year, because costs rise everywhere and the rule cannot tell a saturated audience from an expensive auction.

Run it against history first, look at what it would have done, and set its limits from that rather than from intuition.

Not every area needs to advance slowly. Creative rotation is usually the safest place to grant authority early, because the changes are reversible, their effects are visible quickly, and the cost of a poor choice is a week rather than a season.

Granting it early also solves a real problem in a store, which is that creative fatigue happens faster than an approval cycle. A buyer who can rotate within an approved library, without a separate sign-off each time, keeps the account fresher than one who cannot.

The limit is the library: assets already approved, within the current campaign brief, not a new concept.

Seven checks. One row per area, no blanket permission. Budget authority is bounded by catalogue segment as well as by amount. No permission allows weight to be added where the margin is unknown. Stock cover is a precondition on any increase. Promotion weeks sit outside ordinary permissions and are planned in advance instead. Automated rules have been tested against the last two peaks. And every change made under the authority is recorded, on the day, with its reason.