Answered straight

How to start a digital agency that survives year two

The short answer

Start a digital agency by working out what one delivery person can actually bill — roughly 100 to 115 hours a month once you subtract admin, pitching and rework. Every retainer you sign eats a share of that. Price each account to clear 55–65% gross margin, and hire before utilisation passes 80%.

Updated 26 July 2026 · Written by the Last Agency team · See what SEO actually costs

The short version

  • A digital agency is a capacity business. Everything that goes wrong in year two traces back to selling capacity you didn't have.
  • 176 gross hours a month per head. Around 100–115 of them are billable. Plan against the second number.
  • Target 55–65% gross margin per retainer. Below 40% you can't afford a second person, which means you can't afford to grow.
  • SEO and performance marketing run on different clocks. Staffing them with the same person degrades both.

Service mix: what genuinely sells together

Not every combination of services is a package. Some pair naturally because they share a buyer, a reporting cycle and a set of assets. Others just share a slide.

Three combinations work reliably, and each one exists for a structural reason rather than a marketing one.

  • Search + content + organic social. One buyer (usually the founder or the marketing lead), one set of assets, one compounding timeline. The content produced for search feeds the social calendar, so the marginal cost of the second service is genuinely lower. We package this as the Organic Growth Engine at ₹99,000/mo.
  • Organic social + paid media. Shares creative production and the same weekly rhythm. Organic tests what resonates cheaply, paid scales whatever survives. Ours is the Full-Funnel Starter at ₹75,000/mo.
  • Search + paid media alone is the awkward pair. Same channel family, completely different clocks, and no shared asset except the landing page. It's sellable, but you're running two unrelated programmes under one invoice.

Billable hours per head — the number nobody calculates

Here's the maths every agency owner should be able to do from memory, and most can't.

A full-time person works roughly 22 days a month at 8 hours: 176 gross hours. From that, subtract internal meetings, client calls that overrun, proposals, hiring, training, tool admin and the rework that follows every round of client feedback. Realistic utilisation for a delivery person in a small agency lands somewhere around 55–65%.

That gives you roughly 100 to 115 genuinely billable hours per head per month. Not 176. If you priced your retainers assuming 176, every account you sell is quietly losing money, and you'll discover it around the time you try to hire the second person.

Where a delivery person's 176 monthly hours actually go.
CategoryTypical shareBillable?
Client delivery work55–65%Yes. This is the only bucket the client pays for.
Client calls, updates, reporting walkthroughs10–15%Partly. Budget it inside the retainer or it eats margin invisibly.
Internal meetings, admin, tool wrangling10%No.
Pitching and proposals5–10%No — and it rises sharply the month a client churns.
Training, learning, algorithm updates5%No, and cutting it is how agencies become obsolete quietly.

Gross margin per retainer, in rupees

Gross margin per account is delivery cost against retainer value. Get the delivery cost right and everything else — whether to take a client, whether to hire, whether to raise prices — becomes arithmetic instead of anxiety.

Fully-loaded cost is not salary. Budget roughly 1.25–1.4× gross salary once you add employer statutory contributions, laptop, seat, tool licences and the person's share of software you pay for whether they use it or not.

Work it through with a hypothetical. If someone's fully-loaded cost is ₹65,000 a month and they bill 110 hours, your cost per billable hour is about ₹590. A ₹75,000 retainer that consumes 45 delivery hours costs you roughly ₹26,500 to deliver — about 65% gross margin. The same retainer at 70 hours costs ₹41,300, or 45% margin, and suddenly you can't afford the account manager it needs.

  1. Work out your true cost per billable hour. Fully-loaded monthly cost ÷ realistic billable hours. Do this per role, not as one blended number.
  2. Estimate delivery hours per retainer before you quote. Not after. Track actuals for three months and you'll find you were 30–50% optimistic, like everyone is.
  3. Target 55–65% gross margin per account. That's what funds sales, admin, tools, the founder's salary and the months when a client leaves.
  4. Kill or reprice anything under 40%. A low-margin account isn't a stepping stone to a better one. It's the account that stops you from serving the better one.

The client-count ceiling, and what to do when you hit it

Capacity is the ceiling, and it arrives earlier than the revenue chart suggests. At 110 billable hours a head and 40 hours a retainer, one delivery person carries between two and three accounts properly. Four is possible for a month. Five is a resignation letter with a lag.

When you reach it there are exactly three moves, and only one of them is bad.

Hire. The right call if pipeline is real and you have the cash. Hire when utilisation crosses roughly 80% sustained — not 100%, because a new person takes six to twelve weeks to reach full output and you need the overlap.

Refuse the work. Genuinely fine, and it's what we do. We cap at three new clients a month, partly because our guarantee means we carry the risk if a programme underperforms — you can't do that at volume and mean it.

Take it anyway and hope. This is the one that ends agencies. Service quality drops across every existing account, the two clients who were about to renew don't, and you've traded three good relationships for one new logo.

Why performance and SEO can't be the same person

This is where digital agencies quietly break, and it's a staffing problem masquerading as a skills problem.

Paid media runs on a daily loop: check spend, check CPA, pause the loser, ship a new creative, adjust budgets. Miss two days and you've burned real money. SEO runs on a 90-day loop: research, build, publish, wait, measure. Miss two days and nothing happens at all.

Put both on one person and the daily loop always wins, because it's louder and the consequence is immediate. The SEO work becomes whatever's left on Friday afternoon. Six months later the paid account is fine and the organic programme has produced eleven blog posts and no rankings.

They also fail differently. Paid fails visibly and fast — the dashboard goes red. SEO fails invisibly and slowly, and by the time it's obvious you've lost two quarters. A person carrying both will always catch the visible failure and miss the invisible one.

  • Separate the people, even part-time. Two people at 50% each beats one at 100% across both.
  • Separate the reporting cadence. Weekly for paid, monthly for organic. Reporting SEO weekly teaches clients to panic about noise.
  • Separate the goals. Paid targets cost per acquisition this month. Organic targets qualified leads against a frozen baseline over 90 days. Different numbers, different conversations.
  • Be honest with clients about it. If you're staffing both from one head because you're small, say so and price accordingly. Where retainer money actually goes is worth reading before you write that proposal.

What actually destroys margin in year two

Not price. Almost never price. Four things, and all of them are operational.

Scope creep. The extra landing page, the deck for the board meeting, the "quick" competitor analysis. Individually trivial, collectively a whole extra account you're delivering for free. Write the scope down, then refer to it without apology.

Unbilled revision rounds. Two rounds included, third round quoted. Agencies that don't set this find themselves on round six of a blog post that earns them ₹4,000.

Free work in the sales process. A full audit given away to win a pitch is thirty hours of your best person's time with a maybe attached. Charge for the diagnosis, credit it against the first invoice if they sign.

The founder as permanent bottleneck. Every approval, every client call, every fire. This caps the agency at whatever one person can supervise and makes the business unsellable — because you *are* the business.

Related questions.

How many clients can a small digital agency handle?

Roughly two to three retainers per delivery person, assuming around 110 billable hours a month and 40 delivery hours per account. A three-person delivery team caps out near eight to ten accounts before quality slips. Plan hires against that ceiling rather than against a revenue target.

What gross margin should a digital agency target per client?

55–65% gross margin on delivery cost. That's what funds sales, admin, tools and the months a client churns. Below 40% an account can't support the account management it needs, and you'll find yourself subsidising it with the founder's unpaid hours.

Should a new digital agency offer SEO and paid ads together?

Only if you can staff them separately, even part-time. They run on completely different clocks — paid needs daily attention, SEO needs a 90-day arc — and one person carrying both will always prioritise the loud daily loop and quietly starve the slow one.

When should a digital agency make its first hire?

When sustained utilisation crosses about 80%, you hold three months of the new person's fully-loaded cost in the bank, and there's signed work to fill at least half their capacity. Fully-loaded cost is roughly 1.25–1.4× gross salary once statutory contributions, hardware and tools are counted.

How much does it cost to start a digital marketing agency in India?

Setup is cheap — a proprietorship, GST registration and a core tool stack land in the ₹25,000–₹60,000 range for month one. The real cost is runway. Model six months of fixed costs including your own salary, because the gap between doing work and being paid for it runs six to ten weeks.

Is it better to specialise or offer every digital service?

Specialise first. A broad menu on day one means learning every craft on a client's budget, buying five tool stacks, and competing on price because nothing distinguishes you. Add the second service once the first has a repeatable process and known delivery hours.

Last slot's open

Make this the last growth call you book.

Grab the free strategy call and walk away with a 90-day growth plan — hired or not. Or just text us. Either way, you'll know exactly how we'd win.

Guaranteed or it's free · No lock-in · Free strategy call