How Do I Get My Bakery Staff to Sell Custom Orders?

The end-to-end process from walk-in to booked cake

A custom order is not a sale — it is a small project with a discovery step, a quote, a deposit, a production slot, and a pickup. Staff avoid selling them because every one of those steps feels like risk: risk of quoting wrong, risk of promising a date the kitchen cannot hit, risk of a long conversation while five people wait behind the customer. Fix the process first, and the selling behavior follows almost on its own.

Start by writing the actual path a custom order travels in your shop. Most bakeries have something close to this: a customer mentions an occasion, an associate asks two or three qualifying questions, someone captures the details on a form or in the POS, a price is quoted from a tiered sheet, a deposit is taken, the order lands on the production calendar, the kitchen confirms feasibility, and the customer picks up. Nine steps, and in most shops only three of them are written down anywhere.

The bottleneck is almost always step two and step four. Associates do not know which questions to ask, and they do not know what to charge. Both are solvable in an afternoon. Build a five-question intake card — occasion, date, headcount, flavor, and any dietary restriction — and laminate it next to the register. Build a tiered price sheet that covers eighty percent of your requests: a quarter sheet at one price, a half sheet at another, a two-tier round at a third, with a small set of named add-ons for fondant, hand-piping, or a photo transfer. Anything outside those tiers gets a callback from the head baker within four business hours. Now the associate is never guessing, and the "long conversation" collapses to about ninety seconds.

The other structural fix is the handoff. If the associate has to own the quote from first question to final pickup, they will duck the conversation on a busy Saturday. Split the job: the counter captures the intake and takes a deposit if the request fits a standard tier, and a designated order lead — often the shift manager or the head baker — handles anything custom enough to need a real quote. Staff will sell what they can complete in the time they have.

Where custom orders create or leak revenue

The reason to push custom orders is margin and predictability, not just ticket size. A case pastry is an impulse buy against inventory you already baked and may have to discard. A booked custom order is a pre-sold unit with a known pickup date, a deposit already collected, and ingredients purchased against a confirmed sale. Shrink on a booked order is close to zero. That is the same structural advantage a RevOps team chases when it pushes a sales floor from transactional deals toward contracted, forecastable revenue — the mechanics are different, the economics are identical.

Leak one is the unasked question. Most walk-in customers who eventually order a cake somewhere never got asked in your shop. They mentioned a graduation while paying for two croissants and the associate said "congratulations" instead of "are you doing anything for it?" That single missed question is the largest revenue leak in a retail bakery, and it is invisible because nothing shows up in the POS. You cannot see a sale that was never attempted, which is exactly why the scorecard needs a *quotes attempted* line, not only a *orders booked* line. Score the behavior you can observe, not just the outcome you can bank.

Leak two is quote-to-book fallout. An associate takes fifteen minutes of intake, promises a callback, and the callback never happens or arrives two days later after the customer has ordered elsewhere. Every quote that does not convert is worse than never quoting: you spent labor and lost the customer's trust. Set a hard callback window and track it. If your quote-to-book rate is under fifty percent, the problem is usually response time, not price.

Leak three is underpricing by fear. New associates quote low because they are afraid to lose the order, and there is no reference sheet to protect them. Fondant work, hand-lettering, and multi-tier construction carry real labor hours that a flat per-slice price does not cover. Publish add-on pricing so nobody has to improvise, and audit a sample of booked orders monthly against actual production hours.

Leak four is the missing follow-up. A customer who ordered a birthday cake in March is a near-certain buyer in March of the following year, and a corporate catering client who ordered once for a quarterly meeting will order again in ninety days. Most bakeries capture the phone number and never use it. An occasion list with dated reminders is the cheapest repeat-revenue engine a bakery has, and it converts far better than any general newsletter because the timing is inherently right.

Leak five is the capacity mismatch. Selling custom aggressively without checking kitchen throughput produces late orders, apologies, and refunds that erase the margin advantage entirely. Cap bookings per day by production hours available, publish that cap where the counter can see it, and treat a sold-out Saturday as a success signal to raise price, not a reason to squeeze one more cake in.

Concrete numbers and benchmarks worth tracking

Vague encouragement — "sell more cakes" — produces nothing. Give the counter numbers they can hit in a shift, and give yourself numbers that tell you whether the program is working.

Ask rate. The percentage of transactions where an associate asked an occasion question. Practically, you measure this by observation and self-report rather than the POS. A reasonable starting target for a busy retail counter is asking on one in four transactions — pushing for every single transaction on a Saturday morning rush is unrealistic and will just teach staff to fake the number. Ask on the transactions with signal: a customer buying a dozen, a customer mentioning a date, a customer who lingers at the cake case.

Quotes per associate per week. Set this against your traffic. A single-location shop with moderate weekday foot traffic might target three to five quotes per full-time associate per week when the program starts, then adjust once you have four weeks of baseline. The point of the first month is to establish the baseline honestly, not to hit an invented target.

Quote-to-book conversion. This is your quality signal. Track it per associate. A sharp spread — one person converting well above the counter average, another far below — tells you exactly who to shadow and who to have run the training. If everyone's rate is low, the problem is your pricing, your callback speed, or your lead time, not your staff.

Deposit rate. The percentage of booked orders where a deposit was collected. Aim for near-total. A common structure is fifty percent at booking for standard tiers, with the balance at pickup. No-deposit orders are the ones that get abandoned, and each abandonment is a full production cost with zero revenue.

Average custom ticket versus average walk-up ticket. Compare them directly. The multiple varies enormously by shop, but knowing your own ratio tells you what a single additional booked order per associate per week is actually worth in monthly revenue — and that is the number that justifies whatever bonus pool you fund.

Lead time distribution. Track how many days ahead orders are booked. A distribution clustered inside forty-eight hours means you are constantly firefighting; a healthy program has most orders booked five days or more out, which lets you buy ingredients smarter and schedule labor against known demand.

Repeat rate on custom customers. Measure the share of custom-order customers who order a second time within twelve months. This is the metric that separates a bakery running a real occasion program from one taking orders reactively.

On weighting: pick three to five KPIs, not nine. A workable starting split gives custom quotes booked the heaviest weight — somewhere around forty percent of the composite — with the remainder split across attach rate, basket size, and a service or accuracy measure. Score each line one to five, multiply by weight, sum to a composite. Publish the whole matrix. The exact weights matter far less than the fact that everyone can see them and that they do not change every week.

Pitfalls that quietly kill custom-order programs

Bonusing on booked revenue alone. If the only thing that pays is a booked order, associates will chase large orders and ignore the small ones, and they will get demoralized fast because booking is partly luck of who walks in. Pay on the behavior *and* the outcome — a smaller component for quotes attempted, a larger one for orders booked. Behavior is controllable; outcome is not entirely.

Letting one person hoard the orders. In many shops a single confident associate takes every custom conversation, and the rest of the counter never learns. The composite scorecard exposes this immediately: one person at level five, everyone else at level one. The fix is rotation — the strong associate runs a fifteen-minute session on how they open the conversation, and every associate is expected to own at least a couple of quotes a week.

Scripts that sound like scripts. "Would you like to hear about our custom cake options today?" is a question customers reflexively decline. Teach an opening that references what is already in front of you: "Is the dozen for something special?" or "Are you celebrating?" It is a normal human question and it converts at a dramatically different rate than a pitch.

Skipping the kitchen conversation. Rolling out a custom-order push without the head baker's buy-in produces immediate friction, because the person absorbing the extra work had no say in the target. Set the weights *with* the head baker and the counter lead, and give the kitchen an explicit veto on daily capacity.

Changing the weights constantly. Re-weighting overnight for a holiday push is a real advantage — but only if the baseline is stable the rest of the year. Weights that move weekly teach staff that the scorecard is arbitrary, and they will stop optimizing for it. Change weights at season boundaries or program launches, and hold them at least four weeks.

Measuring nothing for the first month. Owners often set targets on day one with no baseline, then declare failure or success against a number they invented. Spend the first three or four weeks measuring current-state ask rate, quote volume, and conversion. Then set targets a modest step above baseline.

Ignoring the staff who genuinely cannot sell. Some associates are excellent at speed and accuracy and terrible at conversation. Rather than forcing everyone into the same profile, let the matrix carry a service and accuracy line with real weight, and let that associate be the one who handles the rush while a colleague runs the order desk. A weighted scorecard is supposed to reflect a complete counter, not clone one personality.

Neglecting the corporate side. Retail birthday cakes get all the attention, but standing office orders, breakfast platters, and event catering are usually the higher-value, more repeatable revenue. An associate who books one recurring weekly office order has done more for the year than a dozen one-off cakes. Weight catering and standing orders separately so the effort shows up.

Selection checklist for the tooling and the bonus structure

You do not need software to run this. A spreadsheet with KPI rows, weight columns, and a composite formula works for a single location, and it is free and fully transparent. The reason shops move off spreadsheets is upkeep — a sheet nobody updates between seasons is worse than no sheet, because it looks authoritative while being wrong.

If you do buy, decide first where you want the teeth. Order management is one job: bakery-specific systems handle cake orders, deposits, production scheduling, and delivery routing, and they matter when your custom volume genuinely outgrows a paper book. Point of sale is a second job: general retail and restaurant platforms give you per-associate sales data, invoicing, and deposit collection, which are the inputs your scorecard needs. Visibility and gamification is a third: leaderboard and recognition platforms broadcast performance to the floor, which is genuinely effective during a peak season and largely wasted the rest of the year. Compensation is a fourth: commission and attainment tools tie a multi-component score to actual pay.

Most bakeries need the first two and can approximate the last two with a whiteboard and a spreadsheet. Buy the visibility layer only when you have more than one location and cannot walk the floor yourself.

Whatever you pick, insist on three things: you control the weights, associates can see their own scores without asking a manager, and the data comes out in a form you can put in a spreadsheet. A tool that computes a score you cannot inspect will lose the counter's trust the first time someone disputes a number.

Related questions

How long before a custom-order program shows results?

Expect four weeks to establish a baseline and another four to six weeks before booking volume moves meaningfully. Behavior change at a counter is slow because each associate only gets a handful of real opportunities per shift. Judge the program at week eight, not week two.

Should part-time staff be on the same scorecard?

Yes. Score observable behaviors, not tenure or hours. A part-timer working two shifts a week can hit a high ask rate and a strong conversion rate. Normalize volume-based lines per shift worked so the comparison stays fair across schedules.

What if the head baker resists the extra order volume?

Give the kitchen a hard daily cap on custom slots and let them set it. Run a two-week trial with a small bonus pool and compare booked orders against the prior two weeks. Capacity data usually settles the argument faster than persuasion.

Do catering and corporate orders belong on the same matrix?

They belong on the matrix but on their own weighted line. Catering has a longer cycle, higher value, and different skills than a walk-up birthday cake. Blending them into one number hides which associate is actually building repeatable revenue.

How do I keep associates from underpricing custom work?

Publish tiered pricing with named add-ons for fondant, hand-lettering, and multi-tier builds, and route anything outside the tiers to a designated order lead. Audit a sample of booked orders monthly against actual production hours to catch drift.

FAQ

What if my staff is already overwhelmed with counter sales?

Custom orders do not have to add time at the register if the handoff is built properly. Train the counter to capture a five-question intake and hand the customer a one-page form, then let the order lead or head baker follow up. The associate stays on the register; the quote happens off the clock-critical path. Ninety seconds of intake is very different from a fifteen-minute design consultation, and conflating the two is why staff avoid the conversation entirely.

How do I set the weights without confusing my team?

Start with three KPIs and nothing more: custom orders quoted, attach rate on retail add-ons, and basket size. Weight custom orders heaviest — roughly forty percent — and split the rest. Print the matrix on one page and post it in the break room. Add lines only after a month, once the first three are producing honest numbers and people trust the math.

What if an associate scores high on case sales but low on custom orders?

That is precisely the pattern the composite is built to surface, and it is common. Their overall score lands low despite strong till numbers, which makes the gap impossible to argue with. Handle it as coaching, not discipline: shadow them for a shift, identify whether the block is the ask, the pricing, or the fear of the follow-up, and fix that specific step.

How often should I update the scorecard?

Update weights at season boundaries or when you launch a program — holiday cakes, a wedding push, a new catering menu. Hold the same KPIs for at least four weeks so staff can build habits against them. Update the *scores* weekly so the feedback loop stays tight; it is the weights that should stay stable, not the numbers.

Does any of this apply outside a bakery?

Directly. The same structure — weight the behaviors, score them one to five, publish the matrix, tie the bonus to the composite — runs the same way in a florist, a butcher counter, a print shop, or a furniture floor. Anywhere a transactional counter also carries a higher-margin custom or special-order line, the RevOps pattern is identical: make the neglected line visible and paid.

What is the single highest-leverage change if I only do one thing?

Publish a tiered price sheet. Most counters do not decline to sell custom work because they lack motivation; they decline because they do not know what to charge and are afraid of quoting wrong in front of a customer. Remove the guessing and a meaningful share of the reluctance disappears before you have changed a single incentive.

Sources

• https://www.sba.gov/business-guide/manage-your-business/stay-legally-compliant — Small Business Administration guidance on operating and managing a small business.
• https://www.bls.gov/ooh/food-preparation-and-serving/bakers.htm — Bureau of Labor Statistics occupational outlook for bakers, including employment and wage data.
• https://squareup.com/us/en/townsquare — Square's small-business resource library covering POS reporting, invoicing, and deposits.
• https://www.shopify.com/blog/retail-kpis — Shopify's overview of retail KPIs including basket size and attach rate.
• https://www.score.org/resource-library — SCORE's free small-business resource library on pricing, staffing, and operations.
• https://hbr.org/topic/subject/compensation — Harvard Business Review's collection on compensation and incentive design.
• https://www.nrf.com/research — National Retail Federation research on retail operations and consumer behavior.
• https://www.fda.gov/food/food-labeling-nutrition/food-allergies — FDA guidance on food allergen disclosure, relevant to custom-order intake.

Related on PULSE

• [How Do I Know Where, When, and How Many People to Schedule at Each of My Multi-Unit Retail Locations?](/knowledge/tl0001)
• [How Do I Figure Out How Many People to Schedule Each Day and at What Times for My Single Store?](/knowledge/tl0002)
• [How Do I Know How Many Cooks and Servers to Schedule Each Shift at My Pizza Restaurant?](/knowledge/tl0003)
• [How Many Salespeople Should I Schedule Each Day on My Furniture Store Floor?](/knowledge/tl0004)
• [How Do I Decide How Many Reps to Schedule at Each Store in My Mattress Retail Chain?](/knowledge/tl0005)