Social Media Strategy That Decides Where Not to Be

Should we be on every platform?

No. Almost nobody should. Presence on a platform your audience does not use for your category is spend with no upside, and it dilutes the attention the platforms you do need would benefit from.

The reason businesses end up everywhere is that each individual decision seems cheap. Setting up an account costs nothing. Maintaining six accounts properly costs a great deal, and maintaining them badly is worse than not having them — an abandoned profile with three-year-old posts is visible evidence of neglect.

The useful question is not where your audience has accounts, but where they are receptive to a business like yours. Those are different, and the second one is much shorter.

How do we decide what to post about?

From what your business genuinely knows that customers do not. Content pillars built from real expertise sustain; pillars built from what competitors post run out within a quarter.

The test is whether you could talk about a theme for a year without repeating yourself or reaching for filler. Most businesses have two or three of those and mistakenly plan around six.

The other input is what customers actually ask. Questions your sales team answers weekly are proven demand, they are specific, and answering them publicly does double duty as search content as well.

What should we measure?

Whatever maps to the objective you set, which for most businesses is not follower count.

If the objective is leads, measure profile visits, link clicks and inquiries attributed to social. If it is credibility, measure whether people arriving from other channels check your profiles and what they find. If it is retention, measure engagement from existing customers rather than reach among strangers.

What to be sceptical of is reporting built entirely on metrics that only go up. Impressions, follower count and total engagement rise with activity regardless of whether the activity works. They belong in a report as context, not as the headline.

How long before a social strategy shows results?

Organic social compounds slowly and unevenly. Paid distribution shows results quickly. Which one you are relying on largely determines the honest answer, and we will not give you a fixed timeline because it would be a guess.

What is reasonable to expect early is directional signal — which content themes earn attention, which platforms respond, which formats work. That informs the next quarter and is genuinely useful, even though it is not yet a business outcome.

What is not reasonable is expecting organic reach to substitute for paid distribution. Where the objective needs volume soon, the strategy should say so and budget for paid social rather than hoping organic covers it.

How do you decide which platforms to leave?

By being honest about capacity, and by treating an abandoned account as worse than no account.

Most businesses are on more platforms than they can serve, because joining is free and leaving feels like retreat. The result is several accounts that each get a fraction of the attention needed to work, and a visitor who finds the neglected one forms an impression from that.

The decision should follow two things: where your customers actually are, and how much you can genuinely produce. If the honest answer is one platform properly, that is the strategy — and it will outperform four maintained badly.

Leaving does not have to mean deleting. A profile can be kept with accurate information and a pinned post pointing to where you are active. That preserves the search result and the verification value without pretending to a presence you are not maintaining.

What are content pillars and why do they help?

A small set of recurring subjects the account returns to, so that planning becomes selection rather than invention.

The problem they solve is the blank page. Without pillars, every planning session starts from nothing, which is slow, produces inconsistent output, and is the point at which most social programs quietly stop.

Good pillars come from the intersection of what you know, what your audience needs, and what you can produce repeatedly. Three or four is usually right — enough for variety, few enough to be memorable without a document.

They also make the account legible. Someone who sees three of your posts should be able to say what the account is about. Accounts without pillars produce individually reasonable posts that add up to no impression at all.

What should social actually be measured on?

Whatever the objective was, which is why naming the objective first is the whole exercise.

If social exists so that people who look you up find an active, credible account, then reach and follower growth are largely irrelevant. What matters is that the account is current, the information is right, and the impression is good.

If it exists to distribute content, the measure is traffic to that content and what people do when they arrive — which is measured in your own analytics rather than in platform reporting.

If it exists to generate inquiries directly, the measure is inquiries, tracked through to whether they were any good. Platform-reported conversions and your own records will disagree, and yours are the ones that matter.

What is almost never the right measure is follower count, because it can be increased in ways that make everything else worse.

How long before a strategy shows anything?

Longer than the enthusiasm usually lasts, which is the real reason most social programs fail.

We will not put a number on it. It depends on the platform, the starting point, the category, how much is published and how good it is — and any agency quoting a specific timeframe is describing a hope rather than a projection.

What can be said is what to watch early. Whether the schedule is being met, whether the content is getting made without heroics, and whether the response is coming from people who resemble your customers. Those are visible within weeks and they predict whether the rest will follow.

What should not be watched weekly is follower count. Short-term movement is noise, and reacting to it is how a considered strategy becomes a series of reactions to numbers that meant nothing.

How should social fit with everything else you do?

As a distribution and verification layer rather than a separate program, which is how it stops competing with your other channels for content that could serve both.

The most efficient arrangement is that substantial work lives on your own site and social carries people to it. A guide published once can supply weeks of social content, and the traffic lands somewhere you own rather than on a platform that can change its distribution overnight.

It also feeds paid. Organic posting shows which arguments land before money is spent on them, which makes social advertising meaningfully cheaper than starting from an untested creative concept.

And it supports search indirectly. Social links themselves do not function as ranking signals in the way ordinary links do, but content that circulates gets seen by people who cite it, and that is where the effect actually occurs.

What does not work is treating social as a separate content operation with its own subjects and its own calendar. That doubles the production load and produces two bodies of work that reinforce nothing.

How do you set expectations internally?

By agreeing what social is for before anyone reports on it, because most disappointment is a measurement disagreement rather than a performance one.

The common failure is that the team running social is working toward one objective and the person funding it has a different one in mind. Content is judged on inquiries it was never designed to produce, and the program is canceled while doing exactly what it set out to do.

Naming the objective settles it. If social exists so that people who look you up find an active, credible account, then follower growth is not the measure and nobody should be reporting it as one.

It also helps to say what social will not do. It will not replace search demand that does not exist. It will not produce inquiries next month at a predictable rate. Saying so early is less costly than discovering it in a review meeting.

And the reporting rhythm should match the timescale. Weekly reporting on a channel that moves over months produces noise, and reacting to noise is how a considered strategy becomes a series of reactions.

What does a realistic first ninety days look like?

Mostly setup and establishing a rhythm, with very little that looks like results — which is worth saying plainly before it is being judged.

The first part is decisions: which platforms, what the pillars are, who owns it, and what the schedule genuinely is given the capacity available. Getting those wrong is what causes programs to stop, so the time spent there is not preamble.

The second is the profile work nobody counts — information corrected, links fixed, highlights and pinned posts set, and the basic questions answered somewhere findable. It has no visible output and it is what a visitor actually encounters.

The third is building a backlog, so the schedule survives the first busy period. A program that launches with two weeks of content and nothing behind it is a program with a known end date.

What can reasonably be observed by the end: whether the schedule was met, whether producing content is sustainable without heroics, and whether the response is coming from people who resemble your customers. Those predict everything that follows.