Measuring social media success without drowning in metrics

Let us imagine a business with ten employees, such as a landscape gardener, a gym or a dental practice. There is an Instagram and Facebook profile, a Meta ad runs now and then, a small Google Ads campaign looks for customers in the region, and the Google Business Profile is maintained. Each of these places delivers reports. The platforms show reach, impressions, clicks, interactions, profile views, calls and route requests, and every number has its own definition.

The owner usually asks a single question: is this doing anything? No report answers that question on its own. Anyone who lays the reports side by side finds rising and falling curves, but no connection to the order book. Either the evaluation is then forgotten altogether, or people go by the number that looks biggest.

This concept study describes how we would put such a situation in order. It contains no client measurements and no promises about reach or enquiries. It shows an approach: select a few numbers, record them cleanly, read them at a fixed rhythm and derive small, verifiable decisions from them.

One basic idea up front: measuring is not an end in itself and not a control instrument for the agency. It serves to make better decisions with the business's scarce budget and scarce time. If a number does not prepare a decision, it does not belong in the monthly review.

Reach and interactions are the numbers every platform shows prominently. They are available immediately, they move daily, and they feel like success. A post with many likes is a pleasure, and that is legitimate. The problem arises when this pleasure becomes a steering metric. An entertaining post reaches people who would never hire a landscape gardener, and still delivers nice numbers.

In addition, the platforms count differently. Meta distinguishes between reach and impressions, Google between impressions, clicks and interactions, and the business profile counts views in Search and Maps. A number from Meta therefore cannot be compared with a number from Google, even if both have the same name. Anyone who adds them up is calculating with quantities that do not fit together.

Then there is the gap between ad and enquiry. Many businesses are contacted by phone, via WhatsApp or in the shop. These routes appear in no report. Anyone who only measures what the platform counts by itself systematically underestimates the effect, and anyone who then cuts budget may be cutting exactly where enquiries originated.

The delay is also easily overlooked. An enquiry rarely arises on the day of the post. Someone who sees a terrace in February may call in April. Monthly figures have to take this into account, and short-term spikes should be read with caution.

Before we touch a dashboard, we clarify three things. First: what should the channel do for the business? New enquiries, appointment bookings, job applications, local awareness or nurturing existing customers are different goals and call for different numbers. Second: how have people got in touch so far, and which of these routes can be recorded? Third: how much time is available for evaluation? For most small businesses, that is fifteen to thirty minutes a month.

The stocktake produces a list of existing data sources: the channel statistics, the ad accounts at Meta and Google, the business profile, the contact form on the website, the phone, and the question in conversation about how someone found out about the business. We also check whether the accounts belong to the business and who has access. Accounts and data should stay with the client so that evaluations are preserved even after a change of provider.

An uncomfortable truth often emerges: the data basis for the most important question is missing. If nobody notes where enquiries come from, there is nothing to evaluate, however nice the reporting. Then the work begins not with a dashboard but with simple recording. That is less glamorous, but it is the lever with the best ratio of effort to insight.

Our basic idea is restraint. Instead of watching all available metrics, we choose a small group that together tells a story: how many people get in touch, where do they come from, what does it cost us, and how many of them become customers? A small business rarely needs more than that in daily life. Everything else serves troubleshooting and is consulted only as needed.

The four numbers are enquiries per month, enquiries by source, cost per enquiry for paid advertising and the share of enquiries that become jobs. The business supplies the last number itself, because only it knows which enquiry became a job. It is the most valuable because it measures quality rather than quantity. Fifty enquiries with two jobs are worth less than twenty with ten.

Platform numbers such as reach, clicks or ad prices remain important, but in a second row. They help to understand why one of the four numbers has changed. If the cost per enquiry rises, we look into the ad reports for causes. If enquiries fail to come, we check reach and click-through rate. This way the evaluation leads from the decision to the cause and not the other way round.

We also state what we do not promise. There is no commitment that a particular number will rise. Comparison with other businesses is also tricky, because industry, region and season strongly shape the values. More important is the development of your own business over several months.

The most important building block is an enquiry list. This can be a spreadsheet or a field in the existing customer software. For each enquiry, the date, route, source and outcome are recorded. The team asks for the source in conversation or via a selection field in the contact form: Instagram, Facebook, Google Search, Google Maps, referral, other. This is not an exact science, but it provides a picture that platform reports alone do not show.

For paid advertising, technical tracking is added. In Meta ads and Google Ads, goals can be defined, such as submitting the contact form or clicking the phone number. Such goals are called conversions. We would deliberately keep them few: two or three goals that actually correspond to an enquiry, not ten events that sound nice. The prerequisite is a privacy-compliant integration with consent, which the business clarifies together with its data protection adviser. We make no legal assessments in this regard.

Recording also includes the question of phone calls. The phone number in the Google profile provides call numbers, clicks on the number can be counted on the website, and in conversation a sentence like "How did you hear about us?" helps. Many teams find this tiresome at first. After a few weeks it becomes a habit, and the answers are often more revealing than any statistic.

Finally, we define names and meaning of the numbers. What counts as an enquiry, what as a qualified enquiry, what as a job? If these terms are understood differently within the team, the result is numbers that nobody can trace. Half a page of definitions prevents this.

In daily life, evaluation only works if it is small and regular. We would suggest a fixed routine: once a month, always in the same order. First the enquiry list, then the cost per enquiry for paid advertising, then a quick look at profile calls and route requests, and last the platform numbers as an aid to explanation. The result fits on one page and consists of numbers, one sentence of assessment and one planned change.

The change is the actual purpose. Each month the business decides one thing: test a different ad creative, shift a campaign period, take up a topic more often in the editorial plan or adjust the opening hours in the profile. More than one change at a time blurs the effect. The following month, we check whether the number has moved, with due caution, because season and chance play along.

As for budget: the ad money goes directly to Meta and Google, and the business keeps control of the account and payment method. We read the numbers along with you and give recommendations, but switch nothing on without approval. Posts and ads also go to the business for approval before publication. This protects against surprises and makes clear who is responsible.

Anyone who does not want to take over the routine themselves can receive a short summary, in plain language and without a jumble of jargon. What matters is that it stays within the owner's reach. A report that only the agency understands creates dependency instead of clarity.

For small businesses, the numbers are often small. If six enquiries come in per month, the jump to eight is not a trend but noise. We would therefore think in quarters and read monthly values only in context. Cost per enquiry also fluctuates strongly with few data points. It helps to combine periods before drawing conclusions, and not to be guided by individual outliers.

Seasonal businesses need a different yardstick. A heating engineer has more enquiries in autumn than in summer, an ice cream parlour the reverse. Comparison with the previous month is then misleading; comparison with the same month of the previous year is more meaningful, if it exists. In the first year, the only honest statement is that the data is still too thin for conclusions.

A third special case is routes without a trace. Someone who gets to know the business through social media and later types the address directly into the browser shows up as direct traffic. Someone who sees a post and asks in the shop does not show up at all. Here the only help is to ask the person, and we say openly that attribution remains incomplete. A number with known fuzziness is better than a number with false precision.

Finally, there are goals that are hard to quantify, such as awareness or trust. Here we would not invent a pseudo-metric. Instead, one can work qualitatively: what follow-up questions come in, what do new customers say, what do they refer to? Such hints are not statistics, but they are real and often more useful than a curve.

A clean evaluation answers three questions: are enquiries coming in, from where, and at what price? It also shows whether a change is going in the right direction. That is not little, because many businesses have so far had only assumptions. Over time, a picture emerges of which topics, channels and ads work for your own business, and the budget can be shifted on sound grounds.

It is equally clear what it cannot do. It guarantees no enquiries and no reach, and it proves no causality. If more calls come in after a campaign, the campaign may be the cause, but so may the weather, a newspaper article or the season. We would phrase such connections as an indication and not as proof, and we would not base decisions of greater consequence on a single month.

Expectations about time frames should remain realistic. The recording is in place after a few weeks, reliable patterns appear after a quarter at the earliest, seasonal statements after a year. Anyone who demands certainty sooner gets either false precision or disappointment. An interim status with open points is more honest than a report that explains everything.

The most important learning effect concerns attitude. Anyone who measures to confirm themselves will always find a matching number. Anyone who measures to decide needs few numbers and the willingness to correct expectations. This applies to agencies as well as businesses. An honest evaluation may also conclude that a channel is not worthwhile for this business.

A second point: recording in-house is more valuable than any reporting tool. The enquiry list belongs to the business, it is preserved when the provider changes and it connects channels that no platform sees together. It costs little and delivers the answers the owner is really looking for.

Third, routine beats perfection. A simple evaluation maintained for twelve months achieves more than an elaborate system that gathers dust after three months. We would therefore rather start with too few numbers and expand later than the other way round.

And finally: measuring and shaping belong together. The numbers give hints, but topics, tone and presentation are decided by the business with its knowledge of customers and location. A good evaluation supports this judgement, it does not replace it.

Anyone who wants to measure social media success does not need a data department, but a clear question, clean recording and the discipline to read a few numbers regularly. Enquiries, their source, the cost per enquiry and the share of jobs are enough in most small businesses to steer budget and topics sensibly.

We would stay open to what cannot be measured and not feign accuracy. Accounts, data and ad budget stay with the business, posts and ads go to approval before publication, and we give no assurances about reach or enquiries.

If you would like to know which numbers really count for your business, we will go through your existing reports and enquiry routes together and name the few measures that can support a decision. Even the finding that you are already watching the right things is part of it.