Pricing a Social Retainer: A Build-Up Model

Retainer underpricing is the default failure in this business, and it happens for a boring reason: most agencies price by looking sideways at competitors instead of building up from what delivery actually costs. This is the build-up model I use, worked line by line with real numbers, plus where AI-assisted delivery genuinely changes the math and where it does not.

Why underpricing is the default failure

Pricing by competitor comparison feels safe and produces the same mistake at scale: if the whole market is underpriced, matching it just makes you underpriced too, with extra confidence that it is fine because everyone else is doing it. Analysis of 50+ agency proposals by SocialRails found small businesses paying an average of $1,200 a month for full-service management, with most retainers landing between $500 and $5,000. That range is wide enough to hide agencies running at a real loss right next to agencies running healthy margins, doing similar work, because the low end of that range rarely covers actual delivery cost once you account for every hour spent.

The fix is not “charge more because you deserve it.” It is building the number from your actual costs first, so you know your floor before you ever look at what a competitor charges.

The build-up model, worked line by line

Start with a loaded hourly rate, meaning your real cost per hour including overhead, not just take-home pay. For a solo operator or small agency, a reasonable loaded rate lands somewhere between $35 and $75 an hour depending on market and experience; use your own number, this range is illustrative, not a claim about what you specifically should charge [VERIFY: illustrative range only, not sourced to a survey].

Next, estimate real hours per month for one client at a defined scope: say, 4 channels, 3 posts a week per channel, light community management, one monthly report. A realistic estimate for that scope: 2 hours a week on content planning and drafting, 1 hour a week on scheduling and posting logistics, 1 hour a week on community management, 2 hours a month on reporting, roughly 14 to 16 hours a month total.

At a $50 loaded rate and 15 hours, delivery cost alone is $750. Add tool costs (the next section), then a margin on top, typically 30 to 50 percent for a service business that wants to reinvest in growth rather than just cover payroll. That build-up, not a glance at what a competitor advertises, is the floor you should not sell below.

Tool costs in the stack

Real numbers, not vague line items. If you run this client through PostSider at the Standard tier ($20/mo for 5 channels) because the scope fits inside it, that is your entire scheduling tool cost for this client, assuming you are not splitting a shared tier’s cost by client count, in which case divide accordingly. A client needing more channels or a team seat pushes you to Team ($35/mo, 10 channels, unlimited seats) or higher, and that line item should move with the client’s actual scope rather than staying fixed regardless of what they need.

Add whatever else sits in your stack: a design tool subscription allocated per client, an analytics or reporting tool if it is separate from your scheduler, any AI writing or image tool cost. Keep this line itemized in your own internal model even if the client never sees it broken out, because it is the number that tells you whether a specific client is actually profitable once tools are accounted for, not just hours.

The AI question: lower costs, same price, better margin

AI-assisted drafting, PostSider’s post checker and caption rewrite among other tools, can meaningfully cut the hours line in the build-up above: less time spent on first-draft writing, faster caption variants per platform, fewer rounds of internal revision before something is client-ready. That is a real cost reduction.

The decision that matters is what you do with it. Passing the full saving through as a lower price trains the client, and the market, to associate AI-assisted delivery with a discount, which is a hard position to walk back later. The better default: keep the price where the build-up model puts it, let the hour reduction show up as improved margin, and use the freed-up time either to take on more clients at the same staffing level or to raise your quality bar within the same retainer. State this to clients plainly if they ask why AI use has not moved the price: the deliverable and its quality bar are what they are paying for, not the specific hours behind it.

Packaging: three tiers that sell

Three tiers, not one flat price, gives a prospect a decision to make between packages instead of a decision to make with you about the price itself, which is a friendlier negotiation for both sides. A workable default: a lighter tier at roughly your floor from the build-up model above (fewer channels, lower posting frequency), a core tier priced with your full margin at the scope most clients actually want, and a premium tier that adds something meaningfully more expensive to deliver, deeper reporting, faster turnaround, more channels, priced with room for the extra delivery cost plus margin on top.

Anchor the premium tier high enough that the core tier looks like the reasonable middle choice, which is standard pricing psychology and works here the same as anywhere else it is used.

Raising prices on existing clients

The build-up model above assumes a new client at today’s costs. An existing retainer priced a year ago is very likely below what the same build-up produces today, since your loaded rate and tool costs both drift upward over time even if the client’s scope has not changed.

Attach any increase to something concrete: a stated annual adjustment written into the original contract, an expanded scope, a new channel added, a new deliverable. An increase with no attached reason reads as opportunism to a client even when the math genuinely justifies it, and the client remembers which explanation you gave far longer than they remember the actual percentage.

Run your own numbers

The model: loaded hourly rate times real hours, plus itemized tool costs, plus a stated margin, checked against market data like the SocialRails figures above only as a sanity check, never as your primary input. That build-up is what tells you your floor. Everything above it, three-tier packaging, an AI-assisted margin improvement, a justified annual increase, is what turns the floor into an actual pricing strategy.

If channel-count math across a growing client roster is part of what is squeezing your margin, the per-channel pricing tax breaks down what agencies actually pay different scheduling tools at 5, 10 and 20 clients, a number that feeds directly into the tool-cost line in the build-up model above. Check your own tool line against PostSider’s pricing if a growing client roster is what is squeezing the math.