Is Social Media Worth It? Measure It One Channel at a Time

The question usually arrives at the end of a long week. You have posted three times, replied to a handful of comments, spent forty minutes trying to make a photograph of a countertop look appealing, and you are wondering whether any of it produced a single customer. Meanwhile somebody in a Facebook group is insisting that if you are not on TikTok you will not exist in two years, and somebody else is insisting that social media is dead and you should be sending emails instead.

Both of those people are talking about social media in the abstract, which is the one way it cannot be judged. Instagram for a wedding cake baker and Instagram for a commercial plumber are not the same activity, and neither is the same as a solicitor answering questions on LinkedIn. The useful question is much smaller and much more answerable: what does each channel send me, how many hours does it take, and would those hours do more somewhere else? That is arithmetic, and you can do it in an afternoon.

Almost every owner who tells me social media does not work for them turns out to mean one platform. Often it is the platform they were told to be on rather than the one their customers use. A freelance designer might find that a single well-argued LinkedIn post brings three enquiries and their Instagram brings compliments from other designers. A plumber might find that a local Facebook group where residents ask for recommendations is worth more than everything else combined, and that their own business page is a graveyard.

So when you sit down to work this out, do not put a tick or a cross next to "social media". Make a short list of the specific places you spend time: your Instagram account, your Facebook page, the two local groups you post in, LinkedIn, whatever else. Four or five lines is normal. Each line gets its own verdict, and the verdicts will not match.

The number you want is the number of people who arrived on your site from each platform, and what they did next. In your analytics this lives under referrers — the plain-English meaning of that word is simply the website a visitor was on immediately before yours. If someone taps a link in your Instagram bio, Instagram is the referrer.

There is an honest catch here, and anyone who tells you otherwise is selling something. Social apps are leaky. A link shared in a WhatsApp message, a Facebook Messenger chat, or a screenshot someone typed out by hand usually arrives with no referrer attached at all, so it lands in the bucket called direct traffic — visits where no source was passed along. People in the trade call that dark social. It means your social channels are almost certainly doing slightly more than the referrer figures show, and that is a reason to treat small differences as noise rather than a reason to give up on measuring.

So read the numbers as a floor, not a ceiling. If Instagram shows twelve visits in a month, the true figure might be fifteen. It is not one hundred and fifty.

You can close most of the measurement gap in ten minutes by tagging the links you post yourself. A tagged link is an ordinary link to your website with a short label added to the end, which your analytics reads and files under the right heading. It survives being copied into a message, which a referrer does not. Our guide to campaign links walks through building them.

Keep the labels boring and consistent, because you will be reading them for years. The link in your Instagram bio gets one label, the link you drop into a local Facebook group gets another, and the link in your email footer gets a third. Then when eleven people reach your booking page this month, you can see which of those three sent them.

One warning worth having: do not tag links that other people will share on your behalf, such as the main URL on your printed leaflets or your Google Business Profile. Tags are for links you place yourself.

Now the other side of the sum, which is the side almost nobody writes down. For a fortnight, note the minutes. Include the parts that do not feel like work: choosing between two photographs, rewriting a caption, reading comments, the twenty minutes lost to scrolling after you posted, the Sunday evening where you did not post but felt guilty about it. Guilt is not billable, but the scrolling is real time.

Most owners I talk to guess low by roughly half. Someone who says "twenty minutes a day" is often nearer forty-five once the invisible parts are added. That matters, because forty-five minutes a day is around five hours a week, and five hours a week is the biggest single expense in most small marketing efforts. It never appears on an invoice, so it never gets questioned.

Here is what the finished sum looks like. These figures are illustrative rather than from any particular business, but the shape of them will be familiar. Say a bakery reviews one quarter and finds that Instagram sent 340 visits, of which 24 reached the enquiry page and 6 sent the form, producing 4 cake orders. The time spent was around four hours a week, so roughly 50 hours across the quarter. That is about 12 hours of work per order.

Same bakery, same quarter: two local Facebook groups where the owner answers the occasional "anyone know a good baker" question. That sent 95 visits, 18 reached the enquiry page, 11 sent the form, and 7 became orders. Time spent, being honest, around 20 minutes a week. Under five hours for the quarter, and under an hour of work per order.

The conclusion is not "quit Instagram". It is that these two things are not remotely equivalent and should not receive equivalent effort. To run this sum you need to know how many visitors reached a meaningful page or sent a form, which means setting up a conversion — a specific action you have decided counts, such as an enquiry form being sent or the phone number being tapped. Without that, you are comparing visit counts, and visit counts are the easiest number in marketing to inflate and the least connected to money. If you would rather see the whole picture at once before you start configuring anything, a free report card on your site will show you which sources are already sending you people who do something.

There are genuine benefits that will never appear in this arithmetic, and pretending otherwise makes the whole exercise less trustworthy. People who message you directly on Instagram instead of using your website. The customer who walks in and says they have been following you for a year. Suppliers who take you more seriously because your page looks alive. The reassurance a stranger gets when they check you exist before ringing, which is a real job even though it produces no click.

That last one is worth separating out, because it is the strongest honest argument for keeping a channel that sends you almost nothing. A profile that shows recent activity and answers the question "are these people still trading" is doing useful work as a shopfront. But a shopfront needs maintaining, not staffing full time. Posting monthly so the page is not obviously abandoned is a different commitment from posting five times a week.

Do not cut three channels at once, because then you will never know which cut hurt. Take the weakest one on your list, reduce it to the shopfront level for a quarter, and put those hours somewhere specific and measurable. Not "marketing" in general. One thing: rewriting the two pages people actually land on, asking for reviews, sending an email to past customers, or answering questions in the one group that already works.

Then compare the same three months. Total enquiries matters more than traffic, because traffic can fall while enquiries rise, and that is a good quarter. Weekly summaries help here because they surface changes while you can still remember what you did; that is the idea behind our weekly AI reports, which rank what shifted rather than handing you a wall of figures. If you are weighing up whether measuring this properly is itself worth the cost, the plans and what each includes are laid out plainly, and there is a free tier that covers this kind of comparison.

One more thing to check before you decide anything: seasonality. A plumber comparing October to July is measuring the weather, not their marketing. Compare against the same months last year where you can.

In my experience, most small businesses that run this exercise end up with one channel worth real effort, one worth keeping alive at low cost, and one or two they quietly stop. That is a good outcome. It is not a rejection of social media and it is not a conversion to it. It is a reallocation of the only budget you genuinely cannot top up, which is your own week.

The thing to avoid is going another year without doing the sum, because the default answer to "is this worth it" is always "probably, a bit, I think", and that answer costs five hours a week indefinitely. Measure it once properly, decide per channel, and revisit it in six months when your customers have moved on to somewhere else.