How to increase sales in retail: small store experiments

Most advice on ways to increase retail sales is a list of 25 tips, which invites you to try six at once and learn nothing. Treat each idea as an experiment instead. Pick one change, a baseline, a metric, a time window and a decision rule, and write them down before you start.

The retail sales formula

Sales = traffic × conversion × average transaction

• Traffic is the number of people who come in.
• Conversion is the share who buy. Your retail conversion rate is transactions divided by visitors, so 120 sales from 400 visitors is 30%.
• Average transaction value is sales divided by transactions. With 400 visitors, 30% buying and a $25 average, the day brings in $3,000.

Raise one term while the others hold and sales rise in proportion. The four experiments below work on the last two.

Set up every test the same way

1. One change, written as a sentence: "Staff offer a gift bag with every gift sale."
2. A baseline: the same measure for a recent period without the change.
3. A metric the change should move, and one that must not get worse.
4. A time window, fixed in advance.
5. A decision rule, such as keeping the change if units rise by 10% or more and transactions do not fall.

Writing the rule first stops a good week from passing as proof.

Four experiments to run

Fix out-of-stocks on best sellers

For two weeks, check your 20 best sellers at opening and note each day one is empty. Then fix the cause, such as a daily shelf check or a higher reorder point, and measure units sold per day for those 20. Run this test first, because an empty shelf hides the effect of every other change. See keeping shelves stocked and the reorder point calculator.

Move or pair products

Move one steady seller to a busier spot, or place a related item beside it, such as batteries next to flashlights. A product pairing across different categories is called cross merchandising. Count units of both items per day and watch the average transaction. For space measures, see how to measure placement against sales.

Make signs and prices easy to read

In one section, replace missing or handwritten tags with printed labels and add a short sign saying what the product is for. Good retail signage says what it is, what it costs and why it is worth having. Measure units of the signed products per day. The shelf label maker prints tags, and these shelf talker examples show wording. US pricing rules vary by state; NIST's guide to retail pricing laws lists states with item pricing and unit pricing requirements.

Greet and suggest one add-on

Greet everyone soon after they come in. At the counter, offer one add-on that fits the purchase, such as a gift bag with a gift. This is suggestive selling, and one relevant offer sounds like help rather than a script. Count transactions per day for the greeting, and add-on units and the average for the offer. Brief everyone on the rota first.

Measure with the till report

You do not need a people counter. Your till's end-of-day report gives you:

• Units sold per day for the products in the test.
• Transactions, the count of sales.
• Average basket: total sales divided by transactions.

Copy these into a spreadsheet, one row per day. To estimate traffic, tally people through the door for the same hour on the same weekdays. The sales growth tool gives the percentage change.

Avoid fooling yourself

• Compare like weeks: Monday to Saturday against Monday to Saturday.
• Change one thing at a time. If the sign and the display change in the same week, you cannot credit either.
• Note weather, events and promotions. The National Weather Service's climate pages show recorded weather for your area.
• Watch the season. The Census Bureau describes seasonal adjustment as estimating seasonal, trading-day and moving-holiday effects and removing them. A small shop can keep holidays out of both periods instead.
• Keep a comparison: a similar product you did not change. If both rise, the season may explain it.

How long should a sales test run?

At least two full weeks, so every weekday appears twice in each period. Slow sellers need four weeks or more, because a few sales either way swing the percentage. Stop on the date you set, not when the numbers look good.

Worked example: a gift bag offer

This example is fictional. A gift shop open Monday to Saturday offers a $4 gift bag with every gift purchase for 12 open days, against the 12 before. Its rule is set in advance. Keep the offer if bags reach 6 a day and the average transaction rises by $0.40 or more, with transactions not falling.

All three conditions hold, so the shop keeps the offer.

How to calculate sales lift

Subtract the baseline from the test result and divide by the baseline: ($14,994 − $14,400) ÷ $14,400 = about 4.1%. The formula agrees: transactions rose 2% (612 ÷ 600) and the average about 2.1% ($24.50 ÷ $24.00), and 1.02 × 1.0208 is about 1.041.

Then check the story. The 60 extra bags brought in $240, about $0.39 of the $0.50 rise per transaction ($240 ÷ 612). A counter offer does not bring people in, so the owner treats the 12 extra transactions as ordinary variation.

Record every result

Log every test, failures included, so a failed idea is not retried by accident.

Add each change you keep to your daily and weekly checklists.

Where Headoffis fits

Headoffis is not a point of sale. Record sales by hand or import a sales CSV exported from your point of sale, and Sales & stats shows units sold and units per day for each product. Analytics compares the last 7, 30 or 90 days with the preceding period, in sales value and units, not visitors or transactions. A low-stock threshold puts a best seller in Alerts when recorded stock falls to that level. See the analytics help.

← All posts