Suggestive Selling: Techniques That Grow Average Order Value

Suggestive selling is the practice of recommending a complementary item at the moment someone is already buying: the dessert with the dinner order, the vase with the bouquet, the second case of sauce with the weekly wholesale drop. Most of what has been written about it assumes a person standing in front of another person. A server at a table, an associate at a counter, a trained smile and a well-timed sentence.

That assumption no longer describes where most of these orders come from. Among limited-service operators who said off-premises was the more successful side of their business, 54% named delivery as their most important channel, and off-premises accounted for 83% of limited-service customer traffic in 2024 (National Restaurant Association). The suggestion still has to happen. It just has to happen on a phone call, in a text thread, or on a checkout screen that nobody is watching.

There is also a reason to care more about it when you deliver than when you don’t, and it is arithmetic rather than enthusiasm. The cost of getting an order to a doorstep barely moves when the order gets bigger. One extra item on an existing stop is close to pure contribution. This guide covers what suggestive selling is, the per-stop math that makes it worth the effort, the specific phrases and prompts that work on each channel, how to coach a team on them without turning everyone into a telemarketer, and what to watch so you know it is working. The one input it assumes you have is a short list of items worth suggesting, which is easiest to build from the stock your sell-through rate says is moving too slowly.

Key Takeaways

• Suggestive selling recommends a complementary item alongside the purchase; upselling moves the customer to a bigger or pricier version of the same thing. The techniques overlap, the wording does not.
• Cross-selling raises average order value by roughly 10–15%, and bundled offers by 20–30%, with upsell and cross-sell programs lifting revenue 10–30% overall (shno.co).
• When you deliver, the trip cost is already sunk against the first item. An added $9 side on a stop you were making anyway carries almost no incremental delivery cost, which is why a small attach rate moves margin more than it moves revenue.
• Timing beats wording: offers shown after the main item is in the cart consistently outperform offers shown while the customer is still browsing (shno.co).
• Automated add-on prompts raise ticket size by around 3% with no added labor, and online orders already run 15–20% larger than walk-in orders (Paytronix).
• Suggest the items you need to move. Pairing the suggestion list with your slowest-selling stock turns a sales tactic into inventory control.

What suggestive selling is, and how it differs from upselling

Suggestive selling adds something next to the purchase. Upselling replaces the purchase with a larger version of it. “Would you like a dozen cookies instead of six?” is an upsell. “These go well with the cookies, and they keep for a week” is suggestive selling.

The distinction matters because the two land differently. An upsell asks the customer to reconsider a decision they already made, which can read as second-guessing. A suggestive sale leaves the original decision intact and offers to round it out, which is why the recommended item is usually cheaper than the main one and why the ask feels like service rather than pressure. If your team needs the upsell half of the toolkit in detail, that is a separate skill with its own script problems.

Three things have to be true for a suggestion to work:

• It has to be relevant. The item should solve a problem the first item creates. Candles with a cake. Ice with a beverage order. Serving tongs with a catering tray.
• It has to be cheap relative to the order. Add-ons priced at a couple of dollars up to around five clear the mental threshold where people stop doing math.
• It has to arrive after the main decision. Suggest too early and you are competing with the thing the customer came for.

Why an added item is worth more on a delivery order

The economics of suggestive selling change when you are the one driving. Delivery cost attaches to the stop, not to the contents of the box. Whether the order is $60 or $78, the route is the same length, the driver spends the same time at the door, and the fuel burn is identical.

Here is the same add-on on three order sizes, assuming a $12 all-in cost to serve the stop and a 60% gross margin on product:

The add-on raised revenue by 20% and contribution by 36%. That gap is the whole argument. A business that sells on-premises captures the margin on the extra item; a business that delivers captures the margin on the extra item and spreads a fixed trip cost across more of it.

This also tells you where the ceiling is. Suggestive selling cannot fix an order that was never profitable. If the base order does not cover the stop, a $4 add-on will not rescue it. That problem is a minimum order problem, and worth solving first with a storefront that sets delivery zones and order minimums properly before you optimize the suggestion layer on top of it.

Suggestive selling phrases that work on the phone

Phone orders are where suggestive selling still looks like the textbook version, and where most teams quietly skip it. The pattern that works is specific, singular and tied to what the customer just said.

Weak, because it asks the customer to do the work:

“Would you like anything else with that?”

Stronger, because it names one thing and gives a reason:

“That tray feeds about twelve. Most people ordering it for twelve add a second dessert platter so there’s something left at the end — want me to put one in?”

The moves inside that sentence are worth separating out, because they are reusable:

• Name one item, not a category. “Anything else” invites “no.” A named item invites a yes-or-no decision about that item.
• Attach a reason the customer recognizes. Quantity, occasion, what similar customers do, how long it keeps.
• Ask for permission to act, not for a decision. “Want me to add one?” is lighter than “would you like to buy one?”
• Stop after one. Two suggestions on one call reads as a script. One reads as help.

For recurring wholesale and catering accounts, the strongest version of the phrase is historical: “You ran out of the lemon last cycle and ordered mid-week — want me to bump it to three cases so that doesn’t happen again?” Nobody can argue with their own order history.

Suggestive selling at online checkout

Online is where the volume is, and it is the channel where suggestive selling runs without labor. Automated add-on prompts lift ticket size by about 3% and require no staff time at all, and customers ordering online already spend 15–20% more per order than walk-ins (Paytronix). Recommendation engines are doing the same job at enormous scale: roughly 35% of purchases on Amazon come from recommendation algorithms (McKinsey, cited by Envive).

Four placements do most of the work:

• On the product page, tied to the item. “Goes with” beats “you may also like,” because it explains itself. Tie the suggestion to the specific product, not to the whole catalog.
• In the cart, after the main item is in. This is the highest-yield moment. Offers presented once the primary item is in the cart consistently outperform the same offers shown during browsing (shno.co).
• As a bundle rather than a list. “Add a side and a drink for $4” outperforms offering each separately, and bundled offers raise average order value by 20–30% (shno.co).
• On the order confirmation, for the next order. A one-click add before the batch is picked costs nothing and converts on the strength of momentum.

What does not work is a modal that interrupts checkout. Unexpected extra costs already drive roughly 48% of preventable cart abandonments, against an overall abandonment rate of 70.22% (Baymard Institute). Anything at checkout that reads as another charge gets treated like a surprise fee. The suggestion has to be visibly optional, priced in plain sight, and skippable in one click. Getting the mechanics of that checkout right is covered in depth in the guide to building an online store that keeps orders coming in.

Using your delivery minimum as the suggestion

If you run a free-delivery threshold or an order minimum, you already own the single most effective suggestive selling prompt available, and it is one on-premises businesses cannot use at all.

A customer sitting at $42 against a $50 threshold is not being sold to when you tell them so. They are being told how to avoid a fee. The prompt writes itself: “You’re $8 from free delivery — a pint of soup is $8.50.” Stock the threshold gap deliberately with items priced to close it: single desserts, drinks, small sides, a half-dozen of something. If your minimum is $50 and your cheapest add-on is $22, the prompt cannot resolve.

The same logic works in reverse on routes. Two half-sized orders on one street are a worse stop than one full order, so a threshold that nudges each customer up also nudges your route density up.

Which items to suggest, and how to choose them

Most teams pick suggestions by margin. Margin is the wrong first filter, because a high-margin item nobody wants has a zero attach rate and a high-margin item you are about to throw away has a negative cost of inaction.

Build the list from three inputs, in this order:

• What is aging. Stock with a short remaining shelf life or a slowing sell-through should be first in the suggestion queue. This is where sell-through rate tells you what to clear before it ages out. The slowest-moving SKUs in that report are your suggestion list for the week, not a markdown problem.
• What pairs obviously. If the pairing needs explaining, it will not survive a seven-word prompt.
• What travels well. This filter is specific to delivery and it overrides the other two. An item that arrives melted, crushed or separated turns a $6 add-on into a refund, a complaint and a lost reorder. Test every suggestion candidate through a real trip before it goes on the list.

Refresh the list weekly rather than setting it once. A suggestion list that never changes becomes invisible to repeat customers and stops tracking your actual inventory position.

How to coach a team on suggestive selling

Suggestive selling fails in practice for a reason that has nothing to do with the script: people hate feeling like they are pushing something, and when they feel that way they drop the ask. Coaching has to address the discomfort, not just the wording.

What works:

• Give them the sentence, not the principle. “Offer a complementary item” is not something anyone can act on at 7pm on a Friday. “If they order the family tray, say ‘most people add the second dessert platter’ ” is.
• Rotate one suggestion per week. One phrase, one item, everyone. It is learnable in a shift and measurable by Friday.
• Practice out loud before it meets a customer. The ask is a physical skill, and the first ten attempts are always stiff. Role-play is the cheapest way through that, and it is now possible to drill a phone order against a simulator rather than burning a colleague’s shift. Tools like Kendo’s sales role-play let someone run the same objection ten times until the response stops sounding rehearsed.
• Never tie it to individual quotas. Quotas produce the pushy version, which costs you reorders. Track the team’s attach rate, talk about it as a group number.

Set the expectation that most suggestions get declined. Upsell and cross-sell take rates generally land in the 15–30% range (shno.co), so a team member hearing “no thanks” four times in five is performing normally, not failing.

What to measure

Three numbers tell you whether any of this is working, and only one of them is the one most people track.

• Attach rate. The share of orders containing at least one suggested item. This is the operational number, and it reflects whether the ask is actually happening. Measure it by channel, because phone and online will diverge wildly.
• Average order value. The headline number, and the one that moves slowly. Programs of this kind typically lift revenue 10–30% (shno.co), but the early movement shows up in attach rate first.
• Contribution per stop. The number that matters most when you deliver and the one almost nobody has. Revenue per stop minus product cost minus cost to serve. A suggestive selling program can leave average order value flat while raising contribution per stop, if what it did was consolidate two small orders into one bigger one.

Give any change four weeks before judging it. Attach rate is noisy week to week, and a single large catering order can distort average order value enough to hide a real trend.

Where suggestive selling goes wrong

Four failure modes account for most of it.

• Suggesting the same thing to everyone. A regular who has declined the dessert platter eleven times is being told you are not paying attention.
• Suggesting at the wrong moment. Before the main item is chosen, the suggestion competes. After payment, it is friction.
• Suggesting items that do not survive the trip. The margin on the add-on is smaller than the cost of the refund and much smaller than the cost of the customer.
• Treating it as a revenue lever instead of a service one. The suggestion that works is the one the customer is glad you made. Everything else erodes repeat ordering, which is far more valuable than the $6.

If average order value is the actual goal and suggestive selling alone is not getting you there, it sits inside a wider set of effective strategies for increasing sales. Pricing, minimums, bundling and reorder cadence all pull the same lever from different angles.

Frequently asked questions

What is suggestive selling?

Suggestive selling is recommending a complementary product or service at the point of purchase, so the customer buys the thing they came for plus something that goes with it. The recommended item is usually less expensive than the main purchase, and the recommendation is framed as a suggestion the customer can decline without friction.

What is the difference between suggestive selling and upselling?

Upselling moves the customer to a larger, newer or more expensive version of what they are already buying. Suggestive selling leaves the original choice alone and adds a complementary item next to it. A larger size is an upsell; a side dish is a suggestive sale.

What are good suggestive selling phrases?

The effective ones name a single item and attach a reason: “That tray feeds about twelve — most people add a second dessert platter so there’s something left over.” Avoid open-ended questions like “anything else?”, which put the work back on the customer and are easy to decline.

How much can suggestive selling increase average order value?

Cross-selling typically raises average order value by about 10–15%, bundled offers by 20–30%, and overall upsell and cross-sell programs lift revenue by 10–30% (shno.co). Take rates of 15–30% are normal, so most individual suggestions are declined.

Does suggestive selling work on online orders?

Yes, and it needs no staff time. Automated add-on prompts raise ticket size by roughly 3%, and online orders already run 15–20% larger than walk-in orders (Paytronix). The highest-converting placement is in the cart, after the main item has been added.

The bottom line

Suggestive selling is an old technique with a new best use case. For a business that delivers its own orders, every added item rides on a trip whose cost is already committed, which means a modest attach rate produces a margin improvement out of proportion to the revenue it adds.

The implementation is unglamorous. Pick a short list of items that pair obviously, survive the trip and need moving. Give your team one named sentence per week rather than a principle. Put the prompt in the cart rather than at checkout. Use your delivery threshold as the suggestion whenever a customer is close to it. Then watch attach rate by channel and contribution per stop, not average order value alone, because the stop is where your economics actually live.