Networking for small businesses: coffee beats the algorithm

Piazza Adriano, Santa Maria Capua Vetere. It's a Thursday in October, twenty past seven. The light comes in low and creeps across the still-wet paving, a stone's throw from the Anfiteatro Campano. Inside the bar you can hear the sharp knock of the portafilter against the edge of the knock box, and the smell is hot milk and shortcrust pastry. Next to me at the counter is Berto.

He has the hands of thirty years of serving coffee: swollen knuckles, yellow calluses, coffee dust lodged in the lines of his palms. He stirs the sugar into his short espresso at an uneven pace and doesn't look at me. In 2012 he paid an agency three thousand euros for a website that didn't bring him a single extra customer. Ever since, the word "digital" has sounded to him like a con in a tie.

One thousand six hundred kilometres further north, in a taproom in Salford, just off Chapel Street, there's Bobby. The smell is malt and damp wood. Bobby's a brewer, with the same knuckles as Berto and the same look of someone who's learned to distrust headed notepaper. He's nursing a half pint and, if you start on about "automated acquisition funnels", he stares at the ceiling.

Berto and Bobby are archetypes: I've built them by putting together dozens of real faces. They'll never meet, but they buy in the same way. Not from a brand: from the person standing in front of them. They need to look you in the eye, because they want to know whether, when the job goes wrong, you'll pick up the phone or hide behind a ticket.

Berto and Bobby are archetypes. Image generated with AI.

Here's my thesis, and I won't soften it. Face-to-face networking isn't a nostalgic habit for people who can't work social media. It's the only channel where the algorithm can't take your place, and that's exactly why it needs more method than any campaign. An artificial intelligence can write three hundred cold emails in twelve seconds. It can't sweat over an awkward question, it can't pay for the coffee, and it can't shake a hand with just the right pressure.

But there's a flip side, and I'll say it straight away: networking eats a frightening amount of time. Without a method, it's an expensive round of drinks with a business card in your pocket. With a method, you can measure it, improve it and choose. This article is the method.

What the data says, and what it doesn't

I'll start with a confession. I couldn't find equivalent data for Italy, and I have no intention of making it up to make the paragraph look solid. That gap does say something, though: in the UK, networking is an industry with circuits, membership fees and price lists; in Italy, especially down here, we still call it "knowing people" and keep it under wraps, as if measuring it were disrespectful.

The data I do have comes from the Nettl Pulse 2026 report, a survey of 223 British businesses, from micro-businesses to those turning over more than a million pounds. I've read the PDF and I'm giving you the numbers from the "usage versus results" chart. I've added a column of my own, the ratio between the two percentages.

The "yield" column is my own calculation, not a Nettl statistic. In the PDF, the organic social figures are read off the chart and aren't quoted in the text.

Source: Nettl Pulse 2026, 223 British businesses. Graphic rendering generated with AI

Here's what I read in it. Events are the third most used channel and almost level with social for results, but they convert better relative to how many people use them: those who do them seem to get a return. The PDF itself describes them, in a line I like, as a channel that "bypasses the scepticism of digital advertising and builds an immediate rapport".

What these numbers don't say, and you need to know it:

• It's a small sample, and British only. It doesn't apply to Caserta or Nola.
• They're self-reported answers, not measured turnover.
• It's not clear whether the question allowed more than one answer (the "best results" percentages add up to over 100%, so it probably did).
• The yield is a rough index: a rarely used but excellent channel can look better than it really is.

The only other figure I'll cite with confidence is the main barrier for those who don't take part: 25.5% say lack of time. The PDF also lists budget, consistency of content and the difficulty of tracking what works. I'm not quoting the other percentages because I couldn't match them to the labels with any certainty. Better one figure short than one wrong.

And time, I get it completely. When you're the bottleneck in the business, every hour spent holding a glass of warm prosecco is an hour taken off production.

The real problem, as I see it, isn't time. It's confidence in the return. Someone who says "I haven't got time" often means "I don't know if it's worth it". If they were sure of coming back with a five-thousand-euro annual contract, they'd find the time in three seconds flat, even with the Tangenziale at a standstill. And certainty doesn't exist, but an honest estimate does: I'll get to the formula shortly.

First, a word about ragù. Building a business relationship is like making a Neapolitan ragù: you don't lift the lid every thirty seconds to see if the meat has fallen apart. It has to "pippiare", to simmer gently away for hours. Someone who goes to an event, collects three contacts, sends an email the next day and gives up on Friday because nobody's replied is cooking a ragù in the microwave.

Coffee as commercial infrastructure

There's never been an age in which deals were born in offices. They were born where people went to drink, eat and waste time.

In London, Edward Lloyd was running a coffee house in Tower Street as early as 1688. He wasn't a broker: he made coffee. But captains, merchants and anyone looking for someone willing to insure a ship's cargo gathered there. The place moved to Lombard Street in 1691 and the group was formalised in 1871 with the Lloyd's Act. The world's most famous insurance market was born from people looking each other in the eye in a crowded room, not from a corporate strategy.

Illustration generated with AI, not a depiction of a real venue.

In Naples, a century and a half later, the same dynamic had a different aroma. The Gran Caffè Gambrinus opened in via Chiaia, next to Piazza Trieste e Trento, in 1860, and over time became the haunt of artists, writers and intellectuals: the official tourism page mentions D'Annunzio, Oscar Wilde and Matilde Serao. What matters for us is the function. It was a room where you met the right people, and where reputation was built with a greeting, not with a company profile.

Image generated with AI, not a depiction of the Gran Caffè Gambrinus.

Then there's Florence, 12 February 1951. Giovanni Battista Giorgini brings the main American buyers to his residence, Villa Torrigiani, to see an Italian fashion show. From that small room came the system that would later move to the Sala Bianca in Palazzo Pitti. The sources differ on how many buyers were there, so I'm not putting a number on it. What isn't in dispute is the method: invite the right people into a room, let them touch the fabric, get them to have a coffee with the people who make it.

Even companies that seem to have been born digital were born out of the mud of a conversation. In 1998, three English friends took their smoothies to a music festival in London. They asked passers-by whether they should quit their jobs to make them, with two rubbish bins, "yes" and "no", for people to drop their empty cup in. The "yes" bin filled up. The next day all three handed in their notice: Innocent Drinks was born. Their first piece of market research wasn't an online questionnaire, it was a sweaty face at a festival.

The lesson I take from it is simple: the room is never neutral. Whoever knows how to choose the room and read its codes has a disproportionate advantage. And today, when a good share of buying decisions goes through closed conversations (I write about it in Invisible marketing: why customers talk about your brand in DMs and not in the comments), the bar table is one of the few places where you can get into that conversation before the door closes.

Before: choosing where to go

Walking into a room full of strangers with a fistful of business cards isn't networking: it's wandering about. Before you pick your jacket, pick your room. I use four criteria:

1. Match with your ideal customer. Who will actually be there?
2. Attendee list. Can I see it beforehand, or do I have to take the organiser's word for it?
3. Entry fee.
4. Total travel time, including parking, traffic and the unexpected.

The cost-per-qualified-contact formula

The metric that matters isn't what the ticket costs. It's this:

Cost per qualified contact = (entry fee + travel + total hours × your hourly rate + materials) ÷ number of people in your target you actually spoke to

Then you need a second number, the break-even point:

Expected value of a contact = average margin on a job or contract × percentage of contacts who become customers Contacts needed to break even = total cost of the event ÷ expected value of a contact

Two examples with illustrative numbers (typical cases, not real data). I read them with the same grid: first what the evening costs, then what each person you meet is worth, and finally how many you need so you don't lose out.

Bobby, London. He goes to a free networking event in Shoreditch. An hour's travel and two hours of evening: three hours in all. In the room he talks to three independent pub landlords, exactly his target.

Berto, Caserta. He shuts the bar for half the morning and goes to a trade association event in Naples. Two hours of traffic, toll booth and parking, three hours of event: five hours in all. He talks to fifteen people, but they're insurers, estate agents and software salespeople. Not one of them is a possible customer: not one is a manager of a practice or office with at least twenty employees who needs breakfasts and coffee breaks served.

The numbers are assumptions, the method isn't. The point is this: Bobby chose the room knowing who'd be in it, and for him an evening like this, on average, pays. Berto didn't know, and lost 320 euros and half a morning with the bar shut. That doesn't mean London beats Naples: it means neither of them checked the attendee list beforehand.

In England, many circuits charge a fixed fee and have an attendee list you can look at: that's why they filter. Always check the price list on the official site, because it changes.

If you're in London, for instance, The Business Show takes place at ExCeL in November 2026: it's a huge trade-fair event, useful as a barometer but to be tackled with a list of names in your hand, not with idle curiosity.

In Italy the landscape is bumpier. The industrial associations of Caserta and Naples and the Chamber of Commerce put on institutional events of a high standard. And yet, in my experience, the places where a business owner lets their guard down are different: small tables, meetings in company showrooms, evenings organised by a trade association of 40 members. If the local angle interests you, read District marketing: why the lone wolf won't survive in 2026 and No shop is an island: neighbourhood co-marketing.

A word on size. Every digital consultant cites the giant conferences of the big brands as the model. Not me: for a small business, the starting point is the table of twelve. Coffee, as Berto would put it, isn't made with another bar's recipe: you set the grind yourself, every morning, by listening to the machine. The right room is often in a warehouse unit on the main road, not in the conference centre.

Inside: working the room

When you step through the door, time changes speed. Working the room doesn't mean behaving like a shark with flyers: it means having a mental architecture. The linchpin is the thirty-second introduction, the one the American manuals call the elevator pitch and have made insufferably fake. In practice, it's still essential.

It doesn't tell the story of the company since 1980. It answers four questions:

1. Who you work for. The exact target, not "everyone".
2. What problem you solve. Not "we offer solutions": "I help X stop losing money on Y".
3. What proof you've got. A result, a customer you can name.
4. What you're asking for. Who you're looking for in this room.

An example, for a bar owner like Berto. Not like this: "We're a historic bar with quality coffee and fresh croissants for thirty years". Nobody will remember it. Like this: "I work with the professional practices and offices round here that have meetings and breaks to cater. Their problem is that the industrial caterer turns up cold and late. I bring hot breakfasts and coffee to the office by half eight, with no minimum order. Right now I'm looking for managers of practices and offices with more than twenty people." (Example pitch: the details are made up.) It's an assist: you put the listener in a position to score.

In the room, what counts is the balance between listening and talking, not how long you hold the floor. Whoever asks the questions steers the conversation. And the goal, when you leave a huddle, isn't a crumpled business card: it's an appointment, with a date and time, for a phone call. The card is just the paper alibi.

Getting out of a conversation that's holding you hostage is a skill. When the financial adviser explains pension fund rates to you for twenty minutes, he's using up the best part of your evening. I use what I call the service bridge: "What you're saying matters for risk management. Have you got a card? I'd like to think it through properly. Right now I need a quick word with the organiser before it starts, so let's speak Tuesday." A clean cut, polite, and with a date.

When the room punishes you

The room is powerful, in both directions. One wrong joke can cost as much as a lost contract. The textbook example is Gerald Ratner. On 23 April 1991, speaking to an audience of business people at the Institute of Directors, at the Royal Albert Hall in London, he joked about the low price of a sherry decanter: he sold it at that price because, he said, it was "total crap". The audience laughed. The press the next day didn't. In the following months the group lost hundreds of millions of pounds in market value (estimates vary, which is why I'm not giving a single figure), and Ratner left the company in November 1992. The business, which bore his name, was renamed Signet Group (sources give slightly different dates for the name change). The room doesn't forgive arrogance delivered with a smile.

I've had to learn to read the room the hard way myself. October 2019, the Polo della Qualità in Marcianise. Around twenty business owners from the tinned-food sector, people who export tomatoes across half the world and run their factories with an iron fist. I'd been invited to talk about digital positioning. I had clean, perfect, sterile slides. I talked about "omnichannel funnels", marketing automation, lead nurturing. After seven minutes it had gone cold in the room. The silence wasn't attention: it was a ten-second silence that weighed on you, judging. One producer folded his arms and looked at his watch.

I switched the projector off mid-sentence. I said: "These slides are useless to you. Your problem isn't the algorithm: it's the German distributors who knock three per cent a year off your margins, and you haven't got a brand to say no with. Let's talk about that." The room changed temperature in three seconds. If I hadn't been there feeling that silence on my skin, I'd have ploughed on to the end. And at the end of the evening a business owner came over to shake my hand.

That's the difference between a text and a room: a text never tells you when you've lost your audience.

After: follow-up, data, privacy

Italian networking often dies in the boot of the car, that same evening: the jacket on the back seat, the business cards on the floor mat. That's where the return on the event quietly goes out. If working the room is an art, the follow-up is factory work.

A method in three moves

1. Within 24 hours: log it. Not forty-eight, not "Monday". Put each contact into your CRM (customer relationship management, the tool that keeps your contacts on file: it can be a spreadsheet, a free app or a management system) with four mandatory labels:

Add the next step, with a date. If you've got a QR code on your card, point it at a page with UTM parameters (Urchin Tracking Module: three labels, utm_source, utm_medium, utm_campaign, that tell Google Analytics where the visit came from) so you can see which event brings you real traffic. I've written a separate guide on quotes: It's not a document: it's your silent salesman.

2. Within 48 hours: first contact. A short, personal message that keeps the promise. Never a price list.

Example (Italy): "Good morning Marco, it's Berto, we spoke on Thursday in Nola. As promised, I'm sending you the quote for the Monday breakfasts. If you're up for it, let's have a five-minute chat on Tuesday at 11 about your office's coffee breaks."

Example (United Kingdom): "Hi Sarah, great to meet you on Thursday at the taproom night. As promised, here's the link on pricing for craft beer. Free for a quick call Tuesday at 11?"

3. After about ten days: a second contact, with something of value. A useful piece of information, a name to connect them with, an article. Then stop, if there's no reply: pushing further isn't tenacity, it's noise.

For "warm" contacts, WhatsApp Business has become a natural bridge (I talk about it in WhatsApp Business in 2026: the CRM you've already got in your pocket). But email remains the pillar for business relationships.

One last reminder: a contact isn't a customer, and a customer isn't a newsletter subscriber. This is the legal part, and it's where people come unstuck.

Privacy: two countries, two sets of rules

I'm not a lawyer: what follows is a map, not legal advice. For your own case, ask your solicitor or your data protection officer.

The DUAA (Data (Use and Access) Act 2025) is the British law that brought PECR penalties into line with those of the GDPR. The British regulator is the ICO (Information Commissioner's Office); in Italy it's the Garante per la protezione dei dati personali.

What that means in practice, with an example. You met Marco at an evening do, and you promised to send him an article. Writing him a personal message to keep that promise is a one-to-one conversation arising from a real meeting. Adding him to the weekly newsletter going out to the whole company is something else, and in the Italian context you need his consent. The cleanest way is to ask for it: in the second contact, "Would you mind if I added you to the newsletter? It goes out once a month". A yes in a chat is worth more than a thousand loopholes.

A note on "legitimate interest" as a shortcut for mass promotional emails: in the Italian context, I've never seen it stand up against art. 130. Before building a strategy on it, have it assessed by someone who does privacy for a living. The Garante imposed a penalty by order (7 May 2015) over commercial communications sent to recipients collected without consent: it's a precedent to read, not to imitate. For more on the newsletter side, there's also The pixel that watches you read: what's changing in your newsletter from 29 October.

Measuring without kidding yourself

The favourite alibi of people who go to events for nothing is: "networking can't be measured, it's about relationships". That's false. It's a convenient excuse for not admitting you haven't got a method. You can measure almost anything, as long as you pick a few numbers.

A note on that last point. People often say that someone who's looked you in the eye closes sooner and haggles less over the price. It may be true, but I haven't got a source to cite, so I'm not writing it as fact. It's a hypothesis you can test yourself, over twelve months, in your own CRM: compare the length of the cycle by source of contact. If it proves you right, you've got a number of your own, and it's worth more than anyone else's statistics.

There's a limit to measurement, and it has to be accepted. An enormous share of the word of mouth generated in the room travels through closed circuits: voice notes on WhatsApp, emails forwarded between partners, the line said over Sunday lunch ("I needed a photographer, I met a good one on Tuesday, hang on, I'll send you the contact"). I've written about this in The dashboard Berto never looks at (and it costs him dear) and Word of mouth can't be improvised: a guide to referral marketing for small businesses. An obsession with perfect data stops you seeing the underground movement of the market.

Then there's a component that no spreadsheet captures and that in Naples has a historic name: the caffè sospeso. Leaving a coffee paid for a stranger, without expecting anything back. In networking it translates into a specific gesture: introducing two people who can work together, without taking a commission. Creating value without invoicing it straight away builds a debt of gratitude that the market, sooner or later, honours. It's the same principle I describe in Stop selling, start helping.

When not to do it

"Getting yourself seen is always a good thing" is poison for small businesses. Visibility for its own sake is a pastime for people with too much spare time. There are three cases where you shouldn't do networking, and a serious consultant will tell you so even if it costs him an hour's work.

First: you haven't got a clear offer. If you can't explain in thirty seconds what you sell to a ten-year-old, don't set foot in a room. Going along "to see what happens" is fishing without a hook. If you're changing your business model and can't say why anyone should give you money, stay in the shop.

Second: you haven't got the four hours afterwards. If you already know that over the next three days you won't manage to set aside four hours to enter the data, write the messages and make the calls, save yourself the fee and the miles of motorway. Twelve business cards in your jacket pocket until its next trip to the dry cleaner's isn't networking: it's paper collecting at a high price.

Third: your customer isn't in that room. A good tailor doesn't fight the cloth, he works with it. If the audience is made up of tech startuppers who've got nothing to do with you and you sell traditional tax consultancy, don't change the buttons hoping the fabric will adapt. Turn down the commission.

My position is this, and it won't change: two targeted events in a quarter, well prepared and well followed up, are worth more than ten working breakfasts dragged along just to clock in for the sake of appearances. Bad networking is a hidden cost that eats first the owner's energy and then his bottom line, and turns him into a disillusioned man convinced that "nothing works round here anyway".

The handshake that code can't write

It isn't nostalgia. It's a matter of craft. Real networking, rough and imperfect, made of clammy hands, wrong jokes corrected on the spot, coffees that are too strong and glances that cross while you talk about margins, is the workshop with its door open that no algorithm can replicate. While your competitors hand their face over to automated sequences and generated copy, the one who turns up in person takes the territory a metre at a time.

Physical presence doesn't scale. True. And that's exactly why it's worth something.

Don't wait for conditions to be perfect. They won't be. Pick the next event that's right for you, work out what it costs, prepare, and then do the follow-up as if your business depended on it. If you'd like a second pair of eyes before you dive in, drop me a line, from Caserta or from City Road in London: no magic funnel, just half an hour of raw analysis of your real chances. After that, it's your call.