Sell one thing, and make it the thing you can do without help
The instinct is to offer everything: SEO, social, ads, websites, video. It reads as capability. It reads to a buyer as "has done none of these enough times to specialise", and it means every new client teaches you a new craft on their money.
Pick the single service you could deliver, alone, tonight, to a paying client, without subcontracting anything. That's your opening offer. You can add the second service once you've delivered the first one five times and know exactly how many hours it eats.
There's a second reason, and it's the one that matters for cash. A single service means one workflow, one tool subscription, one type of deliverable and one kind of conversation. Five services means five of each before you have five clients.
Where the first ten pitches actually come from
Not from a cold email sequence. Not from a lead magnet. The first ten come from the people who already have a reason to take your call, and there are usually more of them than you think.
Write the list out properly, in this order:
- Lead with a diagnosis, not a deck. Twenty minutes finding three specific problems on their site is worth more than a twelve-slide capabilities document. It also proves you can do the work, which is the only thing they're actually assessing.
- Don't give the fix away entirely. Name the problems and the likely cause. Charge for the plan. A free audit that includes the full remediation list is a proposal you did for nothing — see why free SEO audits are sales pitches.
- Expect a low hit rate and a long lag. Ten conversations might produce two clients, and the second one might sign four months later. That lag is the thing you have to fund.
- Former employers and their competitors. You know their market, their problems and often their numbers. Check your employment contract for a non-solicit clause first, and respect it.
- Ex-colleagues who now run marketing somewhere else. The single warmest source in the list. They've seen you work.
- Vendors and freelancers in your existing network — the developer, the videographer, the PR person. They sit next to buyers who need what you sell and can't provide it themselves.
- Founders in whatever community you're already active in. Alumni groups, industry WhatsApp groups, a local startup meetup. Being present for six months before you sell anything is not wasted time.
- Businesses you personally spend money with. Your dentist, your gym, the restaurant you're a regular at. They have terrible websites and they already recognise your face.
Retainer or project? The answer is cash flow, not preference
Both models are legitimate. They behave completely differently in a bank account, and in year one the bank account is the business.
- Start with retainers where the work genuinely is continuous, and take projects to fill gaps — not the other way around. Retainer vs project has the full comparison.
- If you must take project work, structure it 50% upfront, 50% on delivery, and don't start until the first half clears. This is normal in Indian services and nobody senior will blink at it.
- Three-month minimum on retainers. Anything shorter and marketing gets judged before it has produced anything, which loses you the client and the reference at the same time.
| Retainer | Project | |
|---|---|---|
| When you're paid | Monthly, ideally in advance of the service month | On milestones, or worse, on delivery |
| Revenue predictability | High — you know January's number in December | None. Every month starts at zero |
| Sales effort per rupee | One sale funds many months | One sale funds one project, then you sell again |
| Scope risk | Creeps quietly unless the scope is written down | Creeps loudly, usually in the last week before delivery |
| Best for | Search, social, paid media, anything continuous | Websites, audits, migrations, brand work |
The cash gap that closes agencies, worked out in full
This is the part that gets left out of every "start your own agency" article, and it's the part that ends most of them. Walk through the actual dates.
You deliver a month of work through January. You invoice on 31 January. The client's terms are net-30, so payment is theoretically due 2 March. In practice the invoice waits for a project manager to approve it, then sits in an accounts payable batch that runs fortnightly, then a portion is withheld as TDS. Money lands somewhere in mid-to-late March. Between starting the work and being paid for it: six to ten weeks, and that's with a client who intends to pay you on time.
Now put a salary against it. You hired one person on ₹40,000 a month. Their pay is due on the 1st, every month, regardless. In the same window where you're waiting on January's invoice, you will pay them for February and March. Two salaries out, one invoice in — and that's the healthy version of this story.
Nothing in that sequence involves a bad client, a failed campaign or a mistake on your part. It's just arithmetic, and it's why agencies with a full order book run out of money.
Six ways to close the gap before it closes you
None of these are clever. All of them work, and most new agency owners are too nervous to ask for them in the first year.
- Invoice on day one of the service month, not the last. For a retainer you're selling a period of capacity, not a completed deliverable. Billing in advance is standard, defensible, and takes 30 days out of the gap immediately.
- Put payment terms in the contract, with a late-payment clause. You will rarely enforce it. Its job is to make your invoice the one that gets approved first.
- Ask for an advance on the first month. New client, unproven relationship, both directions — one month upfront is a reasonable thing to request and it funds your onboarding period.
- Chase on day 31, politely and automatically. Not day 45 when you're annoyed. Most late payments in India are administrative, not adversarial, and a reminder to the right person fixes them in a day.
- Keep three months of fixed costs in the bank before your first hire. Salaries, rent, tools. If the reserve isn't there, use a contractor on a per-project basis until it is.
- Cap any single client at roughly 40% of revenue. When your biggest client is 70% of the business, they aren't a client. They're an employer who can fire you by email, without notice pay.
When to stop freelancing and actually register a business
There's no legal moment. There are practical ones, and they arrive in a fairly reliable order.
You're ready when a client asks for a GST invoice and you have to explain why you can't issue one. When you turn down work because you don't have the hours and there's no one to hand it to. When you're spending more than a day a month on admin you'd rather pay someone else to do. And when the same three referral sources have sent you work twice — that's a channel, not luck.
Registering is quick and cheap in India: a proprietorship takes days, GST registration is free on the portal, and you can convert to an LLP or private limited later without drama. The entity mechanics and a month-one cost sheet in rupees are laid out in how to start an SEO agency — the numbers are the same for any services agency, whatever you're selling.
The thing worth registering for isn't legitimacy. It's that a registered business can hire, and a freelancer can only work more hours.