Upselling Secrets: Training Scripts That Don't Sound Pushy

Upselling is not about pressure. It is about guiding somebody to a better version of what they already came in for. Done properly it is indistinguishable from good service, and it can lift spend per head within the first week or two.

Most pub upselling fails for one of two reasons. Either the team was told to "push the specials" without being told which specials actually make money, or they were handed a script that sounds like a call centre. Both feel awful to deliver and worse to receive.

This is the training structure I use behind the bar at The Anchor, and it works hand in hand with the menu engineering that lifts average spend. Short sessions, one focus at a time, and language your team would use with a friend.

Step 0: Decide what you actually want sold

Before any script, work out what is worth pushing. The answer is cash gross profit per serve, not gross profit percentage, because percentage does not go in the bank.

Here is a worked example, and the first move is always to take the VAT out, because the price on the pump includes it. Say a house lager sells at £5.20 and costs you £1.82. Divide £5.20 by 1.2 and the net price is £4.33, so your cash gross profit is £2.51 and your GP is 58%. A craft keg line sells at £6.60 and costs £2.71. That nets to £5.50, giving £2.79 of cash profit at a weaker 50.7% GP. The percentage falls, the cash rises, and every switched pint puts 28p more in your pocket. Run that sum for every line, rank them, and you will often find the "expensive" products are the ones you want in your team's mouth.

Two practical notes when you build the list. Keep it to six lines a week at most, because nobody remembers twelve. And favour draught over packaged where you have the choice: qualifying draught products under 8.5% ABV served from containers of at least 20 litres attract a lower rate of alcohol duty than their packaged equivalents, which protects your margin on exactly the products you most want to move. If your drinks mix has drifted, fix that before you train anyone, using the same approach as rescuing your margins through the drinks mix.

Step 1: Start with questions

Ask one simple preference question before recommending anything. It converts a sale into a conversation, and it tells your team which of the two options to lead with.

• "Do you like crisp or hoppy beers?"
• "Are you in the mood for something sweet or dry?"
• "Something refreshing, or something with a bit more to it?"

Timing is everything. The question has to come before the guest names a drink, not after. Once somebody has said "pint of Carling please" you have missed it, and anything you offer afterwards sounds like a correction.

The mistake to avoid: asking "what can I get you?" and then trying to upgrade the answer. Train the team to open with the preference question the moment they make eye contact.

Step 2: Offer two curated options

Give a clear choice of two. Three or more creates decision paralysis at a busy bar, and one alone feels like a hard sell. Two feels like help, which is exactly what it is, especially when your drinks menu is designed to guide the eye towards the same lines.

Always name the one you want sold second, because that is the one people remember. Keep the price gap small: under about £1.50 on a pint and under about £2 on a glass of wine. A gap wider than that turns a recommendation into an obvious upsell.

Example: "You could go for the house gin, or our local small-batch gin. The local one is a little smoother."

Step 3: Use bridge phrases

Bridge phrases connect the upgrade to the guest's own goal, and they always carry a reason. "It is nicer" is not a reason. "It is fresher because we get it delivered on a Thursday" is.

• "If you want something lighter..."
• "If you are eating, this pairs well with..."
• "If you want to treat yourself..."
• "If you are driving, the alcohol-free one on draught is genuinely good..."
• "If you are sharing, this makes more sense than two glasses..."

The last one matters more than people realise. Wine sold by the carafe must be in 250ml, 500ml, 750ml or one litre, and glasses come in 125ml, 175ml or 250ml, so the carafe conversation is the single easiest way to lift a table's spend without anybody feeling sold to.

Step 4: Know the four moments

Upselling is not one event. There are four windows in a normal service, and most teams only ever use the first.

The check back is the most neglected. Walking past a table with two half-empty glasses and saying nothing is a lost round, every single time.

Step 5: Drill daily

Run a two-minute role play before service. Same format every day so it becomes a habit rather than an event: one product, one preference question, one bridge phrase, one taste.

The taste is not optional. Nobody sells a beer they have never tried, and a team member who can say "I had a half of this on Friday, it is a bit sharper than the house" will outperform a script every time. Rotate a small taster at handover, keep it off the shift itself, and write down what people say about it, because their words make better scripts than yours.

Pick one category a week: beer, wine, spirits, low and no alcohol, food add-ons, coffee and dessert. Six weeks and you have covered the whole bar, then start again.

Step 6: Track and celebrate

Pick one upsell focus each week and measure exactly one number against it. Your EPOS will give you all of these, and if it does not, that is a good reason to revisit what your EPOS system should be doing for you.

• Average spend per head: net sales divided by covers.
• Items per transaction.
• Percentage of spirit serves sold as doubles.
• Share of total pints taken by the target line.

Post the number on the staff noticeboard on a Monday and again on a Sunday. Celebrate wins publicly and by name. Recognition makes the habit stick, and it sits neatly with the other ways to motivate staff without a pay rise.

If you run a cash incentive, fund it from the business and keep it entirely separate from tips. Since October 2024 the Employment (Allocation of Tips) Act 2023 has required qualifying tips and service charges to go to staff in full, with a written policy they can read, so the tip pot is not available to bankroll a sales competition.

Scripts by category

Beer: "We have the house lager, or the local IPA that is fresher and a bit more aromatic."

Wine: "Would you like a glass or a carafe for the table? The carafe works out better per glass."

Spirits: "Single or double?" Ask it as a straight choice, not as "would you like to make that a double?", which invites a no. Remember your house measure must be either 25ml or 35ml, not both, and the measures notice has to be displayed where guests can read it before they order.

Low and no alcohol: "If you are driving, we have the alcohol-free lager on draught, which is a proper pint rather than a bottle."

Food add-on: "Most people add the rosemary fries or the slaw with that."

Coffee and dessert: "Shall I bring the dessert menu, or would coffees be easier?" Offering a choice between two yeses beats offering one yes and one no.

Common mistakes

• Asking "anything else?" without a suggestion. It is a closing question dressed up as an offer, and the honest answer is always no.
• Training once and never repeating. A script used for a week and then dropped is worse than no script, because the team learns that your initiatives do not last.
• Incentivising volume instead of margin. Reward the cash gross profit lines, or you will pay people to sell your cheapest products harder.
• Upselling after the guest has decided. The window closes the moment they name a drink.
• Not backing off. If somebody says no, accept it warmly and immediately. One "no problem at all" protects the relationship for every future round.

Quick checklist

• Cash gross profit calculated for every line, top six chosen for the week.
• One preference question written per category.
• Two-minute pre-shift drill in the diary, same slot every day.
• One metric picked and posted where the team can see it.
• Tasting rota running so nobody sells a product blind.
• Incentive scheme, if you run one, funded separately from tips and written down.

Mini FAQ

How long before it works? Expect movement within a week or two when the scripts are used consistently, and give it a full six-week cycle before you judge it.

What if my team hate selling? Reframe it entirely. Nobody enjoys selling, but most bar staff enjoy knowing more about the products than the guest does. Train product knowledge and the recommendations follow on their own.