Sales coaching

How to read this page

Every number on this page is one of two kinds. Real numbers come from a source we opened, and they name the source and the year. Example numbers are made up to show the math, and they are marked EXAMPLE. An example is not a stat. It is not a promise. It is not what you should expect. Swap in your own numbers as soon as you have them.

The prices in the examples are also just examples. We sell sites at about $1,000, about $500, and a $100 minimum, plus monthly update plans. Those prices flex. We talk it through with each customer to find a fit, so nothing here is a price list.

Why sales is a numbers game, and why that is only half true

Here is the true part. Most people will not buy from you, and most of them will not even answer. That is normal. If you talk to too few people, nothing else you do can save you. A great script read to six people a week still gets you almost nowhere.

It also helps your head. A no is not a verdict on you. It is one result in a long run of results. EXAMPLE: say you close 15% of your conversations. Run the math on a streak: 0.85 multiplied by itself 10 times is about 0.20. So at that example rate, about 1 run in 5 of ten straight conversations ends with zero sales, even if you did nothing wrong. Bad days are built into the math.

Here is the false part. "It's just a numbers game" gets used as an excuse to blast people with junk. The number you send is only one of the numbers in the game. The others are how many answer, how many talk, how many look, and how many buy. Those depend on who you picked, what you said, and whether you are any good to deal with.

Also, each number is a person. Someone had to read that message on a slow day at their shop. Volume with no care costs you. People tell you to stop, they tell their friends, and on a small town block you only get so many first impressions. There are only so many barbers in one city. Burn through the list and the game is over.

So here is how we say it: the numbers decide how many chances you get, and your skill decides what each chance is worth. You have to work both.

Good news for shy reps. Research says people often think their asks will go worse than they do. A 2016 review by Vanessa Bohns found that people who asked strangers for things underestimated how often they would get a yes by 48% on average. Be careful with this one. Those were favors and small requests, not sales. A later study by Bohns and coauthors (2016) found the error shrinks when money is offered, and a sales pitch is about money. So do not read this as "48% more customers." Read it as: your gut is likely too gloomy about asking, so do not let your gut set your number of asks.

The funnel math, one step at a time

A funnel is just a chain of steps. Each step loses some people. Ours has five counts: touches, replies, conversations, previews viewed, and closes.

Say what each word means before you count, or your numbers will lie to you. Here is how we count:

• Touch: one message, call, or visit to one person.
• Reply: a real person answered. Auto-replies do not count. A "no thanks" does count, but track it apart from the interested ones.
• Conversation: a back-and-forth about their business or the site.
• Preview viewed: they really looked at it, either you saw them look or they told you what they saw.
• Close: money received. A "yeah, sounds good" is not a close.

EXAMPLE funnel, with made-up rates: 500 touches at an 8% reply rate gives 40 replies. If 50% of replies become conversations, that is 20 conversations. If 60% of those view the preview, that is 12 previews viewed. If 25% of those buy, that is 3 closes.

Now multiply it all together: 8% x 50% x 60% x 25% = 0.6%. In this example, 1 in about 167 touches ends in a sale. That looks awful. It is also fine, if the sale is worth more than the time you spent. EXAMPLE: at 6 minutes per touch, 167 touches is about 17 hours. A $1,000 sale for 17 hours of outreach is one kind of week. A $100 sale for 17 hours is a very different one. We come back to that below.

The funnel also tells you where to look. If replies are fine and conversations die, your first reply back is the problem. If lots of people view the preview and nobody buys, look at the preview, the price, or the ask. A funnel stops you from "fixing your script" when the script was never the problem.

Which levers move the result most

Here is the part most people miss. The steps multiply, so doubling any one step doubles your closes. Double touches, 3 closes become 6. Double your close rate, 3 become 6. In pure math, every step has the same power.

So the real question is this: which step can you move most cheaply? Here are the same example funnel and some made-up, realistic-sized moves:

• Baseline: 500 touches, 8%, 50%, 60%, 25% = 3.0 closes.
• 100 more touches (500 to 600, up 20%) = 3.6 closes. Cost at 6 minutes each: 10 more hours, every time.
• Reply rate 8% to 10% (up 25%) = 3.75 closes. Cost: a better list and a better first line.
• Conversations 50% to 60% (up 20%) = 3.6 closes. Cost: answering faster and asking a better question back.
• Previews viewed 60% to 75% (up 25%) = 3.75 closes. Cost: showing it live instead of just sending a link.
• Close rate 25% to 35% (up 40%) = 4.2 closes. Cost: practice on the offer and the ask.
• All four rates together, same 500 touches: 500 x 10% x 60% x 75% x 35% = about 7.9 closes.

Three lessons sit inside that list.

First, skill compounds and it keeps paying. In the example, four modest fixes beat a big jump in effort. To get 7.9 closes on the old rates, you would need about 1,313 touches. That is about 81 more hours of outreach every single batch.

Second, the weakest step has the most room. A step at 90% can never even double. It tops out at 100%. A step at 8% could double and still be low. Look for the leak, not the step you enjoy working on.

Third, there are rules on how you move a lever. You can raise a close rate by lying or scaring people. You will just pay it back in refunds, bad reviews, and no referrals. If a lever only moves when you stop being honest, it does not count.

One more lever is outside the funnel: who you pick. A better list raises replies, conversations, and closes all at once, since you are talking to people with a real gap. It is the closest thing to a free lever here. Spend the extra five minutes finding the shop with the dead link instead of messaging the one with a fine site.

Set a numbers goal by working backward from money

Start with a money target. Then walk back up the funnel to find what you need to do this week. "I'll try harder" is not a goal. "65 touches a day" is.

Here is the walk-back, and all of it is an EXAMPLE. The rates are made up: 8% of touches reply, 50% of replies become conversations, 60% of conversations view the preview. Since cheaper sites are usually easier to say yes to, we also made up a different close rate for each price: 25% of viewed previews at about $1,000, 1 in 3 at about $500, and 1 in 2 at $100. We rounded up, since you cannot close half a person.

• EXAMPLE, $1,000 target at about $1,000 a site: 1 close, 4 previews viewed, 7 conversations, 14 replies, 167 touches.
• EXAMPLE, $1,000 target at about $500 a site: 2 closes, 6 previews viewed, 10 conversations, 20 replies, 250 touches.
• EXAMPLE, $1,000 target at $100 a site: 10 closes, 20 previews viewed, 34 conversations, 67 replies, 834 touches.
• EXAMPLE, $3,000 target at about $1,000 a site: 3 closes, 12 previews viewed, 20 conversations, 40 replies, 500 touches.
• EXAMPLE, $3,000 target at about $500 a site: 6 closes, 18 previews viewed, 30 conversations, 60 replies, 750 touches.
• EXAMPLE, $3,000 target at $100 a site: 30 closes, 60 previews viewed, 100 conversations, 200 replies, 2,500 touches.
• EXAMPLE, $5,000 target at about $1,000 a site: 5 closes, 20 previews viewed, 34 conversations, 67 replies, 834 touches.
• EXAMPLE, $5,000 target at about $500 a site: 10 closes, 30 previews viewed, 50 conversations, 100 replies, 1,250 touches.
• EXAMPLE, $5,000 target at $100 a site: 50 closes, 100 previews viewed, 167 conversations, 334 replies, 4,167 touches.
• EXAMPLE, $10,000 target at about $1,000 a site: 10 closes, 40 previews viewed, 67 conversations, 134 replies, 1,667 touches.
• EXAMPLE, $10,000 target at about $500 a site: 20 closes, 60 previews viewed, 100 conversations, 200 replies, 2,500 touches.
• EXAMPLE, $10,000 target at $100 a site: 100 closes, 200 previews viewed, 334 conversations, 667 replies, 8,334 touches.

Look at the $5,000 row. It is the same money, and the work is very different. At $100 a site, you need about 5 times the touches of the $1,000 route, plus 50 sites to deliver in a month. EXAMPLE: at 3 hours to finish each sold site, that is 150 hours of build time, against 15 hours for the 5 bigger sites.

So does the $100 site make no sense? No. It makes no sense as your whole plan. It can make great sense for the right customer: someone who only needs a simple page today, and who may become a monthly customer or send you a friend.

A real month is a mix, so plan one. EXAMPLE: 3 sites at about $1,000, 3 at about $500, and 5 at $100 = $3,000 + $1,500 + $500 = $5,000. That takes 11 closes and about 1,292 touches on the rates above. Over 4 weeks of 5 days each, that is about 65 touches a day. A team of three would each do about 22 a day.

Now the way you set the goal:

• Pick a money target for the month. Make it specific.
• Write the mix of sales you are aiming for. Be honest about what you actually tend to sell.
• Use your own rates once you have them. Use the examples only until you do.
• Divide down to a daily number of touches.
• Add up the hours. Include the time to build previews, take calls, and deliver the sites. If the hours do not fit in your week, the plan is fake. Change the mix, the rates, or the target.
• Check it every Friday. If you did the touches and missed the money, the rates are the problem. If you missed the touches, the activity is the problem. Those need different fixes.

Control what you can. You cannot promise yourself 5 closes this week. You can promise yourself 65 touches a day. Grade yourself on that, and let the money tell you if the rates need work.

How monthly plans stack

Site sales start at zero every month. Monthly update plans carry over. That is why they matter so much.

The amount is not fixed. We work it out with each customer, based on how much upkeep they need. All the math below uses $50 a month as a made-up example amount.

EXAMPLE: you add 6 new monthly customers every month at $50. Some people leave, as they should be able to. Say 5% cancel each month. Here is what happens to the monthly money: month 1 is $300, month 3 is about $856, month 6 is about $1,589, month 12 is about $2,758. You are not selling any harder in month 12 than month 1. If 10% cancel each month instead, month 12 is about $2,153. Keeping people happy is worth real money.

Work the target backward here as well. EXAMPLE: to reach $1,000 a month in plans, you need 40 customers at $25, 20 at $50, or 10 at $100.

You can also look at what one customer is worth over time. EXAMPLE: if 5% cancel each month, a customer stays about 20 months on average (1 divided by 0.05). At $50 a month, that is about $1,000 over their time with you. So a $100 site with a plan that lasts could bring in about $1,100 in total. At 10% cancels, it is about 10 months and $500, for $600 in total. That is why a small first sale is not always a small customer.

Three honest limits:

• A plan has to do real work. List what is included, like menu updates, hours, new photos, and fixes. If you do nothing for the money, you are charging people for forgetting to cancel.
• Make canceling easy, and say up front what happens to the site and the domain if they do. Never hold someone's site hostage.
• Plans use up your time. EXAMPLE: 55 customers at 30 minutes a month is about 28 hours. Count it before you sell it.

Moving between prices without racing to the bottom

Flexible pricing is a strength when the price moves with the work. It becomes a race to the bottom when the price moves because you got nervous.

First, know what a price cut costs. EXAMPLE: a site at $1,000 takes $300 of your time and costs to deliver, leaving $700. Cut the price 20% to $800 and you keep $500. To end up with the same money, you now need 7 sales where 5 used to do it (5 x $700 and 7 x $500 are both $3,500). A cut of 20% on price is 40% more sales. People never feel that when they say "sure, I can do $800."

The rule: the price moves only when something else moves too. What can you trade?

• Scope: fewer pages, or no booking or shop feature.
• Who does the work: they send the photos and the words, and you do not chase them.
• Timing: it goes live when you have an open slot, not next week.
• Changes: one round, not three.
• A monthly plan: less up front, with the upkeep paid over time. Always say the full total. EXAMPLE: "$300 now and $50 a month is $900 over the first year."
• A referral or a real public review, if it is honest and they want to give it. Never make a good review a condition of the deal.

Here is how it can sound:

Ask first: "What do you need this site to do? Calls, bookings, or just looking legit when somebody Googles you?"

Give a range when they ask: "It depends on what you need. Sites like this usually run about $500 to about $1,000. We also have a simple starter from $100. Let me tell you what I'd actually pick for you."

When they ask for the big site at the small price: "Not that version, no. For $500, here's what I can do, and here's what comes out. What would you cut first?"

When they say they only have $400: "OK. At $400 I'd drop [feature] and you'd send me the photos and text. You'd still have [the main thing they asked for]. Does that work?"

When they push below the minimum: "I can't do it for less than $100 and do a good job. I'd rather tell you no than hand you something bad. Your Google profile may be the better free move for now."

When less is the right call: "Honestly, you don't need the $1,000 one. The smaller one does what you told me." That sentence costs you money today, and it is how you get referrals.

Things we do not do: invent a fake "regular price" to discount from, say "today only" when it is not, say "this is the lowest I've ever gone" when it is not, or start at the $100 option just because you are scared to say $1,000. Set your walk-away number before the call, so you are not making it up under pressure.

Follow-up: what the data says, and how to do it

This is one of the few places vendors roughly agree on direction. They do not agree on the details.

Hunter, a company that sells email tools, reported in 2026 on 31 million emails sent by its users in 2025. Sequences of three messages got a 6.8% reply rate, against 3.3% for a single message. It also said people who follow up three or more times get a lower average reply rate. Instantly, another vendor, reported in 2026 that 58% of replies in its 2025 data came from the first email and 42% from follow-ups. Instantly calls 4 to 7 touches the sweet spot. So Hunter says about three and Instantly says four to seven.

Do not over-trust this. Nobody randomly assigned those senders. People who write follow-ups may also be more careful in other ways, so the follow-ups may not cause the whole gap. It is also bulk email, from those tools' own customers, and not one person writing to a barber. But the direction is useful: a lot of replies come after the first try, and the gains fall off fast.

Here is why that makes sense without any stats. People miss things. They were busy on Tuesday. Your second note catches a different moment. That is all a follow-up is. It is not wearing them down.

Rules for a follow-up that respects people:

• Add something new each time. A new detail, a new thought, or a real question. "Just checking in" adds nothing.
• Second note idea: "One thing I forgot to mention: your hours on Google say closed Sunday, but your door says open. I put the door hours on your preview. Which is right?"
• Third note idea: "This is my last note on this. If the site's a no, all good. If it's bad timing, tell me when and I'll check back then."
• Then stop. Our section on scripts has the full rhythm.
• A "stop" ends it at once, whatever step you were on.

EXAMPLE of why you should count people as well as messages: you write to 100 people. 6 reply to the first note. You send a second note to the other 94, and 3 reply. You send a third to the other 91, and 2 reply. That is 11 replies. Per person, that is 11%. Per message sent (285 messages), it is about 3.9%. It is the same work and the same people, with two very different numbers. Remember that the next time you see someone brag about a reply rate.

Timing: less magic than people claim

The internet loves a "best time to send." Here is what we found when we looked, all vendor-reported, all from 2025 data. Belkins, an outreach agency, found Wednesday and Thursday tied as the best weekdays and mornings from 8 a.m. to noon as the best time (0.54%, against 0.40% for late evening and night). It also says that in its past studies evenings were on top, so the answer flipped. Instantly found Tuesday and Wednesday strongest, with Wednesday highest. Hunter listed sending on a specific day of the week among the things you should not expect to move reply rates.

So three vendors give three answers. The gaps are small, and the winner changes. Do not spend your energy here.

Here is the timing that really matters for us, which is our team's judgment and not a study:

• Their day, not the calendar. A cafe at 8 a.m. and a restaurant at 12:30 are slammed. Come in the slow hour. If you do not know when that is, ask.
• Something just changed. They opened a second spot, added a new menu, started a new series, or got a burst of new followers. A note tied to a real change has a real reason to exist. "I saw you opened on Mission" beats any "best time to send."
• Your speed once they answer. When someone replies, get back to them soon, while they are still thinking about it. A 2011 Harvard Business Review piece by Oldroyd, McElheran, and Elkington said most companies are "not responding nearly fast enough" to online inquiries. That study was about web leads, not replies to cold notes, but the same habit is cheap and it is good manners. We left out the big multipliers often quoted from it, since we could not open the full article to check them.
• Your own numbers. After a few hundred touches, you can check which hours get you real talks. Trust that over this page.

Channel choice: count closes per hour

We could not find solid research that says one channel beats the others for selling small websites. Vendor numbers cannot even be compared across channels. Gong, which sells call-recording software, reported in a 2018 article (updated 2026) that the average success rate of connected cold calls was "as low as 1.48%," based on 90,380 calls, where success meant a follow-up meeting that was held. That number starts counting after someone picks up, and it ends at a meeting. An email reply rate starts at send and ends at any reply. Those are different races.

So measure the only number that decides it for you: closes per hour.

EXAMPLE, one month with made-up numbers:

• Email: 200 touches at 2 minutes each is about 6.7 hours. 1 close. That is 0.15 closes an hour.
• Walk-ins: 30 visits at 20 minutes each is 10 hours. 2 closes. That is 0.2 closes an hour.
• Warm intros from past customers: 5 chats at 30 minutes each is 2.5 hours. 1 close. That is 0.4 closes an hour.

Email had the most touches and the lowest result per hour. The warm intro only had 5 chances, so you could not live on it alone, but you should always use up your warm options first. Then fill the rest of your week with the best cold channel.

Our starting guesses, until your own numbers say otherwise:

• Local shops: in person in the slow hour, or a call to the shop line. You can show the preview right on your phone.
• Creators: a real DM or the business email in their bio. Do not dig up personal numbers.
• Anyone you know, or anyone a customer knows: go warm. Ask for the intro.
• Choose the channel they already use. A shop that answers every Instagram comment and never its email is telling you something.

Check the compliance section before you pick. Cold texts are off the table for us. And yes, email rules apply to one note just as they do to a thousand.

Measure your own numbers, because benchmarks are weak

Look up "average cold email reply rate" and you will find answers that do not match. Here are three that we opened, all for 2025 data, and all vendor-reported: Belkins says 0.45% (7.5 million emails from its client campaigns). Instantly says 3.43%. Hunter says 4.5%, and 3% for sales outreach in particular. The top number is about ten times the bottom number, and they are measuring the same year.

Why so far apart? They count differently. Belkins divides by emails sent and throws out auto-replies, and it warns that a campaign can show 5% against people who opened and 0.45% against sends. Instantly also divides by emails sent, but says it counts all replies. Hunter describes its number per recipient. The senders also differ: each company only sees its own customers, and each one sells something to people who read the report. Most blog "stats" you see are copies of copies of these reports.

None of them cover what we do, which is a real person showing a real preview to one barber. A number from someone else's list, offer, and channel cannot tell you if your week was good.

There is another reason not to chase vendor findings: some winners are not honest. Gong's page, which shows no study date or sample size, says "How have you been?" as an opener has a 6.6x success rate, at 10.01%. It also says "Did I catch you at a bad time" makes you 40% less likely to book a meeting. Maybe so. But if you have never met the person, "How have you been?" pretends you know them. We do not use it, no matter what it scores. Test your openers on your own calls, and throw out anything that is untrue, whatever it scores.

Here is how to measure instead. Keep one simple sheet. One row per person, with these columns: date, name, type (local or creator), channel, list source, what you said first, touches, replied, conversation, preview viewed, closed, price, monthly plan, and reason lost. It takes a minute per person. Then each Friday, work out:

• Replies divided by people, and replies divided by touches.
• Conversations divided by replies.
• Previews viewed divided by conversations.
• Closes divided by previews viewed.
• Closes per hour for each channel.
• Money per touch: total money divided by total touches. This one number stops you from fooling yourself.
• The bad numbers too: stops, complaints, and refunds. If your closes are up and your stops are way up, you are not winning.

Watch out for small numbers. They bounce around a lot. EXAMPLE: if your true reply rate is 8%, then 100 touches should get about 8 replies, but simple chance could easily give you anywhere from about 3 to about 13. So one batch at 3% and another at 13% may be the same script. At 400 touches, the expected 32 replies could still swing from about 21 to about 43. Do not throw out a script after 20 messages.

• Change one thing at a time, or you will not know what worked.
• Compare like with like. Barbers and creators get separate rows and separate rates.
• For a first read, wait for about 100 touches, and treat it as a hint.
• Use reasons lost. Ten people saying "already have someone" tells you more than any benchmark.

Your numbers are also only true for now. A new list or a new offer means you start counting again.

The second-site pitch: backlinks and SEO in plain words

Someone says, "I already have a website." A second site can still be worth selling, as long as it has its own honest job. This section is what you can say, what you can't, and why.

First, the words:

• SEO means search engine optimization. Google's own starter guide says it is about helping search engines understand your content, and helping people find it and decide whether to visit.
• A backlink is a link from some other website to yours. Google says it "uses links as a signal when determining the relevancy of pages and to find new pages to crawl." So links matter, but only as one signal among others.
• For local shops, Google says local results are mainly based on relevance, distance, and how well known a business is, and that how many websites link to you and how many reviews you have feed into that. It also says, "There's no way to request or pay for a better local ranking on Google." So the free Google Business Profile and real reviews are first in line, before any second site.

What a second site can honestly be:

• A separate offer. The cafe sells catering to offices. That buyer wants a menu for 40 and a quote form, not the breakfast page.
• A landing page for a campaign. One page for the fall special, with a QR code on the flyers, so people land on the right thing.
• A different audience. A barber who also teaches classes. A creator with a shop or a course apart from their main channel.
• A real extra presence. It has its own content, and it links to the main site where that helps a visitor.
• One more way to be found. Another page that could show up in a different search, a link for a bio, a QR code, or an ad. That is "more ways in." It is not a guarantee that anyone comes through them.

What a second site cannot do:

• It cannot promise rankings or traffic. Google says, "No one can guarantee a #1 ranking on Google."
• It is not a magic backlink. Its link to the main site is for people. Google's spam policies define link spam as creating links "primarily for the purpose of manipulating search rankings." A site built mainly to pass a link falls under that.
• It will not work fast or surely. Google says some changes "could take several months."
• A keyword domain will not do it. Google says keywords in a domain name alone have "hardly any effect" beyond breadcrumbs. Do not sell "bestbarberfremont" as the secret.
• It does not fix a weak main site. Sometimes the right answer is a new page on the site they have. If so, tell them, even if it costs you the sale.

Now the risks, stated plainly. These come from Google's spam policies. Google says, "Sites that violate our policies may rank lower in results or not appear in results at all." The same policy covers these, and any of them can hurt your customer:

• Link schemes. Buying or selling links for ranking, and "Excessive link exchanges." Paying for posts with links. It also lists widely spread links in the footers or templates of many sites, so do not stuff keyword links into footers across customer sites.
• Doorway pages. "Doorway abuse is when sites or pages are created to rank for specific, similar search queries." Think 30 near-copies of "plumber in [city]" that all funnel to one site. Do not build that.
• Scaled low-value content. Making lots of pages mainly to game rankings, including with AI, without adding value. Google names "Creating multiple sites with the intent of hiding the scaled nature of the content." A second site full of filler is exactly the thing to avoid.

Google's page about hiring SEOs has one more line you should know: "Be wary of SEO firms and web consultants or agencies that email you out of the blue." That is us, when we cold email. A smart owner may have read it. So never pitch yourself as an SEO wizard. Show the real work, and say what you do not promise.

A simple test before you sell one: if Google did not exist, would this site still have a reason to be here? Would a real customer use it? If yes, sell it. If it only exists to rank, do not.

Do the math in the open too. Let them see what it would take to pay off, and let them use their own numbers.

EXAMPLE: a second site at $500. If a new customer is worth $50 in profit to them, it takes 10 new customers to pay for it. If a new customer is worth $250, it takes 2. Then ask them: "Could this catering page bring in 2 new office orders over the next year? You know your customers better than I do." We make no promise that those customers will come.

Safe ways to say it to a customer:

"Your main site does its job. This would be a separate page just for catering, so office managers land on the right thing. It would link back to your main site, too."

"It gives people one more way to find you, and it gives you one clean link to put on flyers and in your bio. I can't promise it'll rank, and nobody honest can. Google says so themselves."

"Search takes time, and I can't control it. Your reviews and your Google profile matter a lot there. What I can promise is what I'll build, and when."

"If a new page on your current site would do the job, I'll tell you. You might not need a second site at all."

What not to say:

• "This will get you on page one."
• "We guarantee more traffic," or "more customers."
• "A second site gives you a powerful backlink that boosts your main site."
• "We have a special relationship with Google."
• "We'll build you a network of sites."
• "Google loves this," or anything else you cannot back up.
• "Your site is hurting you" when you have not checked.

If you want to know whether it works, measure it. Give the second site its own QR code or link, and ask new customers how they found the business. Tell the customer what you see, good or bad.

Everyday talk: why normal people skills matter here

Most of this page so far has been math. Math tells you how many chances you need. The rest of this page is about what you do with a chance, and about the regular talking you do at the counter, at a meetup, at a kid's game, or in line.

Note what the research below is. Most of it is about normal conversation, in labs and on trains. It is not about sales. We use it as a guide to how people work, not as proof that it will close deals.

Start with some comfort. A 2018 paper by Boothby, Cooney, Sandstrom, and Clark found people systematically underestimate how much the other person liked them and enjoyed talking to them. They call it the liking gap. You walk away replaying the dumb thing you said. They walk away thinking you were fine. So when you feel a chat went badly, you may be wrong.

Openers that do not feel like openers

You do not need a line. You need to notice something and say so.

In a 2014 study, Epley and Schroeder asked commuters on trains and buses to talk to a stranger. They reported a better ride than those who sat alone. Other commuters had predicted the opposite. Part of the reason is that people underestimate how much others want to connect. Not everyone wants to chat, though. Headphones on means no. Respect that.

Openers that work in real life, since they are about what is in front of you:

• "How long has this place been here? I keep walking past it."
• "What's good here? I've never ordered anything but coffee."
• "Is that the new one? How do you like it?"
• "You look like you've done this a time or two. How'd you learn?"
• "I'm [name]. I don't think we've met."

Bad openers are fake or too big. Skip "Has anyone told you that you look like..." and skip any compliment on someone's body. A question about their life story is too much for the first ten seconds.

Then give them an out. Say your thing, and let them choose. If they answer with one word and look away, that is the end of it. Smile and go back to your coffee. No harm done.

Reaching back out counts too. A series of experiments by Liu and coauthors (published online in 2022) found people underestimate how much others appreciate hearing from them, and the gap is bigger with people they are less close to. That was about people in your social circle, not cold prospects. So text the old coworker. Just do not read it as permission to spam strangers.

Listening and asking good questions

Most people are waiting for their turn to talk. Being the one who is not is a real edge.

A 2017 paper by Huang, Yeomans, Brooks, Minson, and Gino looked at live conversations, including speed dates. People who asked more questions were liked more, follow-up questions most of all. Here is the fun part: people did not expect asking questions to make them more liked. So the thing that works is the thing most people forget to do.

A follow-up question proves you heard the last thing they said. It is the easiest skill on this page.

They say: "We've been slammed since the street got fixed."

Weak: "Nice. So do you have a website?"

Better: "Wait, how long was the street torn up?"

They say: "My daughter does the Instagram."

Weak: "Cool. Instagram's big."

Better: "Did she start it, or did you make her?"

Ways to listen better:

• Pause a beat after they stop. People often have a second thought, and that one is the real one.
• Say back a few words: "Torn up for eight months?" They will keep going.
• Do not top their story. When they say their car broke down, do not tell your better car story.
• Do not fix it unless they ask. Sometimes they just want you to hear it.
• Ask about the part they sound excited or annoyed about. Their voice tells you where the good stuff is.
• Questions should be real. If you already know where you want them to end up, that is an interrogation, and people feel it. Ask what you want to know.

Share as well. If you only ask questions, it feels like an interview. Give a bit back: "I worked a counter like this for two years. Lunch rush broke me." Then toss it back to them.

Remembering the details

People notice when you remember them. In three experiments published in 1995, Howard, Gengler, and Jain found that when someone remembered a person's name, that person was more likely to go along with a purchase request. People took it as a compliment. Forgetting the name did not lower compliance. So a name helps, and forgetting one is no disaster. Just ask: "I'm so sorry, I forgot your name." Most people are fine with it.

Here is how to actually do it:

• Say the name back once right away: "Good to meet you, Dana."
• Write it down as soon as you can. Your phone is fine. Do not trust your head.
• Note one real detail, like "daughter runs IG," "knee surgery in March," or "wants a second chair."
• Check your note before you see them again.
• Bring it up in a normal way: "How did the knee end up?"

Keep it to what they told you. Remembering what they told you feels warm. Bringing up things you dug up about their home or their family feels creepy. Never open with something you found on their personal page that they did not share with you for business.

And actually care. If you only ask about the knee because your CRM told you to, people can usually tell. Ask because you are curious how it went.

Humor: use it, but keep it small

Humor helps and humor can hurt. In eight experiments published in 2017, Bitterly, Brooks, and Schweitzer found that jokes that land can raise your status, since people read them as a sign of confidence and competence. Jokes that fail, like inappropriate ones, can lower it. Inappropriate jokes still read as confident, but they also read as low competence, and that mix does damage. The title of the paper is "Risky Business," and that is the right frame.

So here are safe moves:

• Laugh at the situation, or at yourself in a small way: "I've called you three times. At this point I'm basically part of the staff."
• Keep it light and brief, and let them laugh or not.
• If it falls flat, do not explain it. Just move on.
• Be quick to laugh at their jokes when they are actually funny.

Stay away from:

• Jokes about their business, their prices, or their site. They built that.
• Jokes about race, religion, gender, politics, or anyone's body.
• Teasing a person you just met. You have not earned it.
• Sarcasm in text. It reads as rude.
• Jokes when they are stressed, or when something bad just happened.

Being funny is not required. Being easy to be around is. If jokes are not your thing, warmth does the same job.

Reading the moment

Before you say the next thing, ask yourself: do they want more of this right now?

Signs to keep going:

• They ask you a question back.
• Their answers get longer.
• They turn their body toward you or put the phone down.
• They bring up a problem without you asking.

Signs to wrap up:

• One-word answers.
• Eyes moving to the door, the screen, or the customer.
• Keys or phone in their hand.
• "Anyway..." or "Well, it was good to..."
• A customer walks in. Step aside before they ask.

Those are clues, not proof. Someone can be quiet because they are tired or because they are thinking. When you are not sure, ask out loud: "Is this a bad moment? I can catch you later." It is a real question, so take the answer.

This matters more than you would guess, as people are bad at it. A 2021 paper in PNAS by Mastroianni, Gilbert, Cooney, and Wilson looked at 932 conversations. They found conversations "almost never" ended when both people wanted them to, and people had little idea when the other person wanted to stop. So do not assume you know. Give people easy exits, and you take the guessing out of it.

Ending well

Most people end conversations badly. They trail off, or they stand there waiting for somebody to say something. Since people rarely know when the other wants to stop, it is a kindness to end things clearly. And people remember how it ended.

A good ending has three parts: name something from the chat, say what is next (if anything), and leave.

• "I'll let you get back to it. Good luck with the second chair. I hope that guy shows up."
• "I've got to run. I really liked hearing how you got this place. See you around."
• "Thanks for the time. I'll send that link tonight like I said, and if I don't hear back, no worries."

Do what you said you would do. A small promise kept is worth more than a great talk. If you said you would send the name of a good sign guy, send it that night, whether or not they ever buy.

Leave a little early. It is better for them to wish you had stayed than to wonder when you were leaving. And no last-second pitch as you walk out. "Oh, by the way, before I go..." wrecks a good talk.

Turning a normal chat into a business chat without being weird

It is weird when you hide what you are doing, or when you force it. It is fine if it comes up honestly.

The first rule is honesty about why you are there. If you walked in to sell, say so in the first minute. Our scripts already do that. Do not act like a customer or a new friend and then flip into a pitch. People feel tricked, and they are right. The ideas below are for real chats, where you did not come to sell.

The easy path is to let them ask. If you ask good questions, people often ask, "So what do you do?" Have a short true answer ready, then stop talking:

"I build websites for small shops and creators. What got you into this?"

That is it. Say one line, and hand it back. If they want more, they will ask.

If they mention a real problem, ask a real question about it. If it turns out you can help, ask before you shift gears:

They say: "Half my customers still ask if we're open Sundays."

You say: "Where do they usually check?"

They say: "Google, I guess. It's wrong and I don't know how to fix it."

You say: "That part I can actually help with. It's a five-minute fix, and I can show you for free. It's also kind of what I do for work, so tell me if you ever want to talk about that. Otherwise, no pitch."

Then give the help, and really leave the rest alone. A free fix with no strings is the best ad you will ever run.

Ways to ask permission:

• "Can I ask you a work question?"
• "This is kind of my line of work. Do you want my take, or are you just venting?"
• "Want me to send you something on that, or leave it?"
• "I don't want to turn this into a sales thing. Should I give you my card and leave it at that?"

With people you know, be upfront: "Hey, this is a work text. I'm doing websites now. Do you know any shop owner who needs one? Don't feel weird about saying no." Do not bring up old times for three messages to soften them up and then pitch. They will see it, and they will remember it the next time you say hi.

Friends can say no, and it should cost them nothing. If they feel your friendship was the price, you have lost more than you could ever make on the sale.

When not to pitch

Some moments are not sales moments, no matter how good the fit is. Do not pitch when:

• They are grieving, or someone close to them is sick. Say you are sorry and mean it. If it was a business chat, go away. Come back only if it feels right, much later.
• They just told you the business is failing, or they are scared about money. Do not sell them hope. If a free tip helps, give it.
• They seem confused about what they would be agreeing to, or are not able to weigh it. Base this on what they do and say, never on age or how they look. Slow down. Suggest someone they trust look at it, or simply stop.
• They have been drinking, or they are clearly upset.
• They are with their kids, at a funeral, at worship, at a hospital, or in any place they came to be somewhere other than work.
• There is a customer in front of them, or their staff is watching and they would be embarrassed.
• It is a friend's party or a host's event, and you were not asked. You are a guest.
• They already said no, or told you to stop.
• You are talking to a minor. Talk to the adult who runs things, or leave it.

Fear is the worst sale. Saying "you'll go under without this" to a scared owner is pressure on someone who cannot think straight right now. Even if it is true, it is not your line. Also never make up a deadline to speed someone up who is already nervous.

Skipping a pitch is not losing. A person who saw you hold back during a bad week will remember that. Some will call you later. Some never will, and it was still the right thing to do.

Put it to work this week

Here is a plan you can start Monday:

• Start your sheet. Write down what each word means, as above.
• Pick one money target. Use the walk-back to get a daily touch number. Mark it as a guess.
• Do the touches. Log them all.
• Pick one lever, the leakiest step, and work on it for the week.
• In regular life, try one thing: ask one real follow-up question in each chat today. Write down one name and one detail.
• On Friday, look at your own numbers, and change one thing.
• After about 100 touches, replace the first example number on this page with your own.

Honest persuasion wins the long game here. You show people something real, tell them what it costs and what it can't do, make it easy to say no, and keep count. The count gets you more chances. How you treat people decides what those chances are worth.

Sources

• (Mis)Understanding Our Influence Over Others: A Review of the Underestimation-of-Compliance Effect (Bohns) · Current Directions in Psychological Science · 2016
• For a dollar, would you...? How (we think) money affects compliance with our requests (Bohns, Newark, Xu) · Organizational Behavior and Human Decision Processes · 2016
• The Liking Gap in Conversations: Do People Like Us More Than We Think? (Boothby, Cooney, Sandstrom, Clark) · Psychological Science · 2018
• Mistakenly Seeking Solitude (Epley, Schroeder), press release · University of Chicago Booth School of Business · 2014
• The Surprise of Reaching Out: Appreciated More than We Think (Liu, Rim, Min, Min), press release · American Psychological Association via EurekAlert · 2022
• It Doesn't Hurt to Ask: Question-Asking Increases Liking (Huang, Yeomans, Brooks, Minson, Gino) · Journal of Personality and Social Psychology · 2017
• What's in a Name? A Complimentary Means of Persuasion (Howard, Gengler, Jain) · Journal of Consumer Research · 1995
• Risky Business: When Humor Increases and Decreases Status (Bitterly, Brooks, Schweitzer) · Journal of Personality and Social Psychology · 2017
• Do conversations end when people want them to? (Mastroianni, Gilbert, Cooney, Wilson), abstract record · PNAS, via Europe PMC · 2021
• Press release on Do conversations end when people want them to? · EurekAlert · 2021
• The Short Life of Online Sales Leads (Oldroyd, McElheran, Elkington), abstract record · Harvard Business Review, via BYU ScholarsArchive · 2011
• 2026 State of Email Outreach (vendor-reported) · Hunter · 2026
• Cold Email Benchmark Report 2026 (vendor-reported) · Instantly · 2026
• Cold email response rates study (vendor-reported) · Belkins · 2026
• Is cold calling dead? (vendor-reported, by Chris Orlob) · Gong · 2018
• Cold Calling quotes and statistics (vendor-reported, no study date shown) · Gong · 2026
• Spam policies for Google web search · Google Search Central · 2026
• SEO Starter Guide · Google Search Central · 2025
• Do you need an SEO? · Google Search Central · 2026
• Link best practices for Google · Google Search Central · 2025
• Tips to improve your local ranking on Google · Google Business Profile Help · 2026