Notes for owners · Industry playbooks
How to advertise a SaaS product
Buy one event, a booked demo or an activated trial, and nothing else. Which channel produces it depends on whether your buyer already knows the category exists.
The GullySales team · Updated 15 Sept 2026 · 7 min read
Advertise a SaaS product against one event, not against impressions. Decide whether your win is a booked demo or a trial that reaches the step where the software does its job, then buy only what produces that event. Buyers who already know the category come from search, because they type the problem and the word software together. Buyers who have never heard of the category need writing and video before an advertisement does anything. And if nobody speaks to a new trial user by day two, more traffic will not change the paid account count.
Pick the event you are paying for
Sign-ups are not the thing. A sign-up with no data in it is a cost.
Pick the first moment where the product has clearly helped, and make that your target. For a billing product it might be the first invoice created. For a field service product it might be the first job assigned to a technician. In a school ERP, it is the first fee receipt issued. Every rupee of advertising is judged against how many of those happened, and at what cost.
This one decision changes the whole campaign. Keywords that bring students and job seekers get cut. Landing pages stop shouting about features and start showing the screen where the work happens. The sales team stops chasing people who typed an email address into a form and left.
Does your buyer know the category exists?
There are only two kinds of SaaS buyer, and they need opposite treatment.
The first already knows. A distributor in Hubli types "GST billing software for distributors" because his auditor told him to get one. A chit fund company searches for exactly that phrase. This buyer is comparing three products on price, support, and whether anyone local uses it. You reach him with search advertising, with pages that answer the comparison directly, and with a demo that can happen today.
The second does not know. A workshop owner losing money because job cards are on paper is not searching for software. He is searching for nothing at all. Reaching him means writing about the job card problem, making short videos in the language he speaks, and letting an advertisement carry that piece rather than the product page. This is slower and it is the only thing that works.
Most Indian SaaS companies sell to both and run one campaign for both. That is why the numbers look bad.
Where the money goes, by buyer state
| Buyer state | What to buy | What to count |
|---|---|---|
| Knows the category, searching now | Search ads on problem terms and competitor comparisons | Demos booked, demos held |
| Knows the category, not searching yet | Comparison pages, feature pages, integration pages | Organic demo requests per month |
| Feels the problem, no idea software exists | Short video, the owner's own writing, WhatsApp and community groups | Shares, replies, calls that mention the piece |
| Already signed up and gone quiet | Onboarding calls, messages, a human by name | Trials reaching the first real use |
| Paying, but shallow | Training, feature adoption, quarterly check-ins | Renewal rate, seats added |
The bottom two rows are not usually called advertising, and they decide whether the top rows pay for themselves.
The trial that dies on day two
Here is where most Indian SaaS budgets leak. A trial starts on Monday. The user opens it, finds an empty screen, does not import a customer list, and never comes back. The campaign report still counts a sign-up.
Fix it with people, not features. A message within an hour from a named person, on WhatsApp, offering to import the first fifty customers or products. A twenty-minute call on day two with the screen shared. One check on day five. For example, a company selling a ₹1,500 a month billing product finds that trials with an import call convert several times better than those without. That changes what the advertising is allowed to spend.
Do that before increasing the budget. Doubling traffic into a broken onboarding doubles the cost of the same number of paying accounts.
What rarely pays back
Advertising in English to a buyer who runs his business in Kannada, Tamil or Hindi. The product screen may be in English. The advertisement and the demo should not have to be.
Long forms. Asking company size, industry and city before a demo request loses the people who were ready.
Paying for clicks on generic terms such as "software for business". They bring students, resellers and people in other countries.
And the free forever plan used as a marketing strategy with no path to a conversation. It fills a database and tells you nothing about who would have paid.
What to do next
Open last month's numbers. Write down three figures: sign-ups, how many reached your first real use step, and how many paid. If the gap between the first two is large, the next quarter's work is onboarding, not advertising. If the gap is small and the volume is low, the work is search and the pages that answer comparison questions. If you would like those numbers taken apart with you, book the free audit.