How Digital Agencies Add Recurring Revenue with WordPress

Why project revenue puts a ceiling on your agency

Project work resets to zero on the first of every month. You can have the best quarter in your history and start the next one needing to win everything again.

That is survivable. What it costs you is less obvious: it makes every other decision worse. You cannot hire confidently against revenue you have not won. You take the client you would rather decline because March looks thin. You discount in December to fill January.

A HubSpot survey found 57 percent of marketing agencies hold at least three months of available cash flow, which is a polite way of saying 43 percent do not. Those are the agencies making decisions from a position of mild panic, and it shows in who they hire and what they agree to.

Recurring revenue does not make an agency bigger. It makes it calmer, and calm agencies make better calls.

What agency recurring revenue actually looks like on WordPress

The WordPress sites you have already shipped are the inventory. Every one needs updates, backups, monitoring, security patching, performance work and small changes forever, and most of them are currently getting that attention either free or not at all.

The streams worth building, roughly in order of how easy they are to start:

Care and maintenance plans. Updates, backups, uptime monitoring, security, a small monthly allowance of changes. The foundation, and the thing every other stream attaches to.

Performance and Core Web Vitals retainers. Sold separately, at a higher rate, because the outcome is measurable and the client watches the number move.

Content and conversion work. Landing pages, A/B tests, new templates. Development capacity sold as an ongoing allocation rather than a project.

Access-based subscriptions. The client pays monthly for priority access to your team with no fixed deliverable list. Highest margin, hardest to sell, and only works once trust is established.

The margin ladder: why selling access beats selling hours

A pattern sits underneath those four, and it is worth internalising before you price anything.

Reported target gross margins climb as you move away from selling time. One agency accounting firm puts retainers at 40 percent and up, managed services at 65 percent and up, and subscription contracts at 75 percent and up, specifically where you supply access rather than hours.

Treat those as directional rather than gospel, since they come from a single source. The direction is the useful part. Once your plan is denominated in hours, you have capped your margin and invited a conversation about whether the client used them all. Sell an outcome and the hours become your business. Sell access and the client stops counting entirely.

What to put in a WordPress care plan at each tier

Most agencies design three tiers before they have sold one. Build one plan, sell it twenty times, then split it based on what people actually ask for.

When you do tier, the split that works is not about how much maintenance you do. It is about how much of your people’s attention the client is buying.

What to include in an entry-level WordPress care plan

Updates, backups, uptime monitoring, security scanning, an SSL check and a quarterly report. Automated end to end, with a human glancing at the dashboard once a week.

If this tier takes more than thirty minutes per site per month, your tooling is wrong, not your pricing. Somewhere between $99 and $199 a month is the common band, and the point of this tier is not the revenue. It is that it gets you a signed agreement and a monthly touchpoint with a client who would otherwise disappear for two years.

Keep it genuinely thin. The temptation is to add “a bit of support” to make it feel worth the money, and that one phrase is how an entry tier becomes a negative-margin obligation.

The mid tier where care plan revenue actually lives

One to two hours of real development time every month, plus everything above, plus priority response and a staging environment for changes.

This is the plan your business runs on. Somewhere in the $400 to $800 range for most small-agency clients, and the reason it works is that the client receives something every month rather than paying to avoid a problem they cannot see.

Above it sits an allocated-capacity tier, often $1,500 a month and up, where the client buys a standing share of your team plus a quarterly roadmap conversation. Fewer clients qualify for it. The ones who do tend to stay for years.

Those bands are starting points, not benchmarks. Your market, your client size and your cost of delivery all move them, and the section below on capacity will move them more than anything a competitor charges.

Why reselling hosting is the worst recurring revenue you can build

Search this topic and you will be told to resell managed hosting within about four results. Notice who is telling you.

The WordPress-specific articles ranking for agency recurring revenue are published, almost without exception, by hosting companies. One lists reselling hosting as strategy number two and follows it with a subheading explaining why their own product is the turnkey way to do it. That is not advice. That is a sales page wearing a blog post.

The advice is bad for four reasons, and none of them are about margin.

You absorb all the liability for infrastructure you do not control. When the site goes down at 2am, your phone rings, not theirs, and your ability to fix it depends on someone else’s support queue.

The price is public. A client can find your host’s retail pricing in thirty seconds and work out your markup, which turns a trusted relationship into an awkward conversation about a $30 line item.

It anchors you as a utility. Once you are the hosting vendor you are something to be switched on price, and that conversation recurs every renewal.

And the support load does not scale with the revenue. One demanding client on a $40 a month plan can eat an hour a week, which is a negative-margin relationship you are contractually obliged to maintain.

The exception, and it matters. Bundling hosting inside a care plan, never itemised, is fine. There it is a cost of delivery rather than a revenue stream, it gives you control of the environment, and the client buys the outcome instead of the server. Same logic as bundling pass-through costs rather than listing them. Sell hosting as a product and you have built the weakest stream available to you. Fold it into something larger and it quietly does useful work.

Why most WordPress care plans fail to sell

Because they are sold as insurance, and nobody enjoys buying insurance.

The pitch is almost always some version of protection. We will keep it updated, we will keep it secure, we will keep it backed up. All true, all valuable, and all completely invisible when you do the job well. A month in which nothing broke feels, to the client, like a month in which nothing happened.

So the renewal becomes a referendum on whether anything bad would have occurred anyway. That is an argument you cannot win, because your evidence is an absence.

The fix is not a better pitch. It is a different product.

Ship something visible every month. A change they asked for. A page speed number that moved. A form that now sends to the right person. A plugin you removed because it was slowing the site down. One concrete thing the client can point at and say that happened because we pay these people.

A monthly report is not that thing. A report is a receipt, and receipts only matter if something was bought. Most agencies send the receipt and skip the purchase, then wonder why the plan churns at renewal.

Build an allowance of real work into every plan and make using it the default rather than a request. The plan survives because the client experiences it.

Recurring revenue is won at launch, not in month six

There is a timing problem underneath every care plan, and it decides the outcome before the pitch does.

Clients reliably say yes at handoff. The site is new, they are slightly anxious about it, and they have just spent a meaningful sum on something they do not know how to maintain. The need is vivid.

Six months later it is invisible. Nothing has broken. The site works. You are now asking them to start paying for a problem they do not believe they have, and the answer is almost always no, or worse, a polite later.

Which means your attach rate is decided by how you run proposals and handoffs, not by how you designed your plans.

Put the care plan in the original proposal, priced, with the first three months included in the project fee and the ongoing fee stated plainly. It stops being a second sale and becomes a line in a document they already agreed to. Agencies that do this consistently attach most of their builds. Agencies that pitch it afterwards attach a fraction, then conclude care plans do not work in their market.

How to move existing clients from projects to recurring contracts

Everything above is easier for new work. The back catalogue is harder, and the usual advice to email everyone is how you get a five percent response and a bruised ego.

Start with your top fifth. The clients who already call you first, whose sites you know well, and who would notice if you stopped answering.

Do not pitch a plan. Pitch the next six months of their website. Go in with two or three specific things you would do, drawn from the site you already maintain informally: the checkout step that loses people, the pages that crawl on mobile, the plugin nobody has updated since 2023. Then say this is the kind of work a monthly arrangement covers.

You are not asking them to buy maintenance. You are showing them a roadmap and offering a way to fund it. Those convert very differently.

Anchor it to something real, too. The emergency you fixed last spring. The migration you handled at short notice. Clients remember being rescued, and a plan is easier to justify when it is framed as the reason the rescue was fast.

The capacity math nobody runs before selling care plans

This is where most agency recurring revenue quietly dies, and it is an arithmetic problem rather than a sales one.

Run the arithmetic before you sell a single plan. Automated maintenance scales beautifully: with decent tooling one person can run updates, backups and monitoring across sixty or eighty sites without strain. That part is fine.

The development allowance is what breaks.

Thirty clients on a plan with one hour of monthly development is thirty hours a month. Fifty clients is fifty hours, close to a third of a full-time month. Sell the two-hour version of that plan, which is the one most agencies actually sell, and fifty clients is a hundred hours. That is more than half a person, permanently committed, before a single project starts.

Those hours do not arrive evenly either. They cluster in the last week of the month when clients remember they have an allowance, which is also the week your project deadlines land.

Why care plan work always loses the internal priority fight

This is the part that actually causes churn, and it has nothing to do with your pricing.

Project work has someone shouting about it. There is a launch date, a client expecting a demo, a deadline with consequences. Care plan work has none of that. The client asked for a small change and will probably wait politely.

So when the week gets tight, and it always gets tight, care plan work is what slips. It slips again the following week. Three months later you have a client who has paid $1,800 and received two small changes and a PDF, and they are not angry, they are just not renewing.

Nobody decided to under-deliver. The capacity simply got spent on whatever was loudest.

Why care plan capacity has to sit outside your project team

The fix is structural. If the same people serve both, project work wins every time, because project work is the thing with a deadline attached.

Either you ring-fence a person’s time so completely that project managers cannot touch it, which is hard at 15 people and impossible at 10, or you source that capacity from outside your project team entirely.

This is the honest reason white-label WordPress development fits recurring revenue so well, and it is a different reason from the usual one. It is not mainly about cost. It is that outsourced capacity cannot be raided by an internal deadline. The hours exist whether or not your own week went sideways, which is precisely what a monthly promise requires.

At Coderlift this is the work we do behind agency care plans: the fixes, the small builds and the improvements that the plan promised, delivered under the agency’s name, while the agency keeps the client, the relationship and the margin.

Guardrails: scope creep in a care plan is a different animal

Project scope creep has an end date. Care plan scope creep does not, which makes it worse.

The mechanism is familiar. Every request is small enough that invoicing for it feels petty, so it gets absorbed, and the absorption becomes the expectation. By month four the client genuinely believes unlimited small changes are what they bought, and they are not being unreasonable. You taught them that.

Five guardrails, written into the agreement before the first invoice.

Define maintenance versus new work in plain language. Fixing a broken form is maintenance. Building a new form is work. Write the distinction down with examples, because arguing it in the moment always costs you.

Cap the rollover. Unused hours either expire monthly or bank for a maximum of two months. Without a cap you eventually face a client with nine stored hours and a project-sized request.

Set response times by severity, not a blanket promise. Site down gets four hours. A broken checkout gets one business day. A copy change gets five. A single blanket SLA either overpromises on the trivial or underpromises on the critical.

Name the exclusions explicitly. New page templates, new third-party integrations, content entry beyond a stated volume, and anything requiring design. These are the four requests that arrive dressed as small ones.

Write the annual increase into the contract. This is the one agencies skip, and it costs more than all the others combined. A plan sold at $600 in 2023 is still $600 today at most agencies, because raising it means an awkward conversation nobody schedules. State a yearly uplift in the original agreement, three to five percent or pegged to inflation, and it becomes a line item rather than a negotiation. Clients who agreed to it on day one almost never query it on day four hundred.

How agencies reduce churn in recurring revenue clients

Churn in care plans is rarely dramatic. It is a budget review in which nobody can remember what the line item does.

Four things move the number, and the first does most of the work.

Deliver something visible monthly, as above. A client who received real work does not cancel it.

Put a name on the account. Not a shared inbox. A person the client can picture, who knows their site and says hello on the quarterly call.

Bill monthly, contract annually. Monthly billing is what clients want, annual terms are what you need, and almost nobody objects when the price is fair.

Have the renewal conversation before they do. Thirty days out, send a short summary of what the year actually produced and ask what next year should focus on. You are reframing a cost review as a planning session, and planning sessions do not end in cancellations.

One honest caveat. A CFO survey found 65 percent of businesses adopting recurring revenue models run into operational problems, and agencies are no exception. Recurring revenue is a promise to do work every month whether or not you have capacity that month, which is a different discipline from project work.

How recurring revenue changes what your agency is worth

Worth knowing even if you never intend to sell.

Buyers value project revenue and recurring revenue differently, and the gap is large. Project revenue depends on a pipeline, relationships and often on the founder personally, none of which reliably transfer. Contracted monthly revenue does transfer, so it carries a materially higher multiple.

The same logic governs anything else that depends on your financial stability. A lender assessing your agency, a landlord on a five-year lease, a senior hire deciding whether to join: all of them read predictable revenue as lower risk.

It also changes how you behave long before any of that. An agency covering its fixed costs from recurring revenue can turn down a bad client, and the ability to say no is worth more than the revenue it costs.

The recurring revenue metrics worth tracking monthly

Five numbers, and the last is the one nobody tracks.

Monthly recurring revenue. The headline. Watch the direction more than the figure.

Churn, measured both ways. Logo churn counts clients lost. Revenue churn counts dollars lost. Losing one large plan and gaining three small ones looks like growth on one and a decline on the other.

Average revenue per account. Rising means you are selling better plans. Falling means you are discounting to close, which is a slower problem but a real one.

Recurring revenue as a share of fixed costs. The threshold that actually matters, covered above.

Hours delivered against hours sold. This is the margin killer and it hides in plain sight. If you sold one hour a month and are delivering two and a half, your 65 percent gross margin is closer to 25 and no dashboard will tell you, because the extra hours were never billed and so never recorded. Track it per client, monthly, and act on the outliers.

When not to sell a WordPress care plan

Some clients should be refused, and knowing which ones protects the whole book.

The client who treats everything as urgent. A care plan is a promise to respond, and someone whose every request is an emergency will consume a tier’s worth of attention at the entry price. Price them at what they will actually cost, and accept it if they decline.

Sites you did not build and cannot assess. Taking on an unknown codebase at a fixed monthly fee is underwriting a risk you have not measured. Do a paid audit first, then decide. Some of those sites should be rebuilt rather than maintained, and finding out in month two is expensive.

Business-critical sites on a budget tier. A store doing real revenue needs a response time and a rollback plan that an entry plan cannot fund. Sell the tier that matches the risk or do not sell one, because the gap between what they paid for and what they expect will surface on the worst possible day.

Frequently asked questions

Which WordPress plugins help agencies manage recurring billing and memberships?

Separate two things that get confused constantly. For billing your own retainers, Stripe Billing or an invoicing tool like Harvest or FreshBooks beats running subscriptions through WordPress, because your revenue belongs somewhere your accountant can reach it. WooCommerce Subscriptions and MemberPress are for building membership sites your clients sell to their own customers, which is a service you deliver rather than a way to get paid. Using a client-facing membership plugin to invoice your retainers is a mess you will regret at year end.

What tools manage updates and backups across multiple WordPress sites?

ManageWP, MainWP and InfiniteWP are the long-standing options for running updates, backups and uptime checks across a portfolio from one dashboard. Most hosts also bundle staging and scheduled backups, which covers part of the job. Choose the dashboard before you sell the plans rather than after, because it is what makes a care plan deliverable at twenty sites instead of five.

How do I handle a care plan client whose site breaks because of something we did not touch?

Fix it, then decide whether it falls inside the plan. A plugin conflict after an update is yours. A third party editing the database at midnight is not, and the agreement should say so. The judgment call is worth making in the client’s favour the first time and in writing the second, because goodwill early in a plan buys more than the hour costs.

How much of an agency’s revenue should be recurring?

Forget percentage targets and find the line where recurring revenue covers your fixed monthly costs: payroll, rent, software. Below it you are selling to survive. Above it every project you win is profit, and the quality of your decisions changes the week you cross it.

Can care plan delivery be automated?

Partly. Updates, backups, uptime monitoring, security scanning and reporting all automate well and should, because they are the low-value half of the work. What cannot be automated is the judgment about whether an update broke something, and the visible monthly improvement that keeps the client subscribed. Automate the maintenance so your people spend their time on the part the client actually notices.