What is digital marketing? A practical guide for Indian businesses

Most explanations of digital marketing are a list of channels. That is accurate and almost useless, because it tells you what the parts are called without telling you what any of them are for, or which one your business needs first.

Here is a more useful definition. Digital marketing is the work of getting the right person to the right page, and then making that page do its job. Everything else (the acronyms, the platforms, the dashboards) is detail underneath those two clauses.

The seven channels, and what each is genuinely good at

Search engine optimisation (SEO). Earning visibility in search results without paying per click. Good at: capturing people who already want what you sell, cheaply and repeatedly, for years. Bad at: speed, and anything nobody searches for.

Paid advertising. Google, Meta, LinkedIn and the rest. Good at: speed, precision and testing. You can learn in three weeks what SEO takes six months to tell you. Bad at: stopping. The moment you switch it off, it is gone.

Organic social media. Good at: building familiarity with people who are not currently buying, which is most of your future customers. Bad at: attribution. It works, and it is very hard to prove in a spreadsheet.

Content marketing. Articles, guides, video, anything that answers a real question. Good at: earning trust before contact, and feeding both search and social. Bad at: producing anything at all if nobody owns the calendar.

Email marketing. Still, by a wide margin, the highest-return channel per rupee for most businesses. Good at: converting people who already know you. Bad at: finding new people.

WhatsApp and messaging. In India this is not a minor channel. It is often the main one. Good at: conversation, follow-up, service and re-engagement, at open rates other channels do not approach. Bad at: scale without discipline, and unforgiving of anything that reads as spam.

Your website. Not usually listed as a channel, which is the mistake. Every other channel ends here. It is the only asset you fully own, and it is where most programmes quietly fail.

The distinction that matters more than the channel list

Split the channels by the job they do.

Demand capture reaches people who are already looking: SEO, Google Business Profile, search ads, directory and comparison listings. The intent already exists; you are competing for it.

Demand creation reaches people who are not looking yet: social, content, video, influencer work, display. You are manufacturing intent that did not previously exist.

Most wasted budget comes from getting this backwards. A business in an established category, a clinic, a builder, a manufacturer, that pours money into content while ignoring the people searching for them today is leaving cheap demand on the table. A genuinely new offering that buys search ads discovers there is nothing to buy, because nobody searches for a thing they have never heard of.

Establish which situation you are in before you decide which platform to use.

The funnel, without the diagram

Someone becomes aware you exist. They consider you against alternatives. They act. Then, if you are good, they come back and tell someone else.

The practical value of that sequence is diagnostic. When something is not working, it is failing at exactly one of those stages, and each has a different fix:

• Nobody knows you exist → a reach problem. More channel, more budget, more distribution.
• People arrive and leave → a message or offer problem. More budget makes it worse, faster.
• People enquire but do not buy → a follow-up or sales problem. Marketing cannot fix this one.
• People buy once and vanish → a retention problem, and by far the most expensive to ignore.

Diagnose the stage before you choose the tactic. Most proposals skip this step, because the diagnosis sometimes concludes that you should spend less.

What has actually changed recently

Three shifts are worth understanding, because they change what good work looks like.

Search results answer the question now. AI summaries at the top of Google, and answer engines like ChatGPT and Perplexity, increasingly resolve a query without the user clicking anything. Ranking is no longer sufficient, you need to be the source the answer is assembled from. In practice that means clear, self-contained answers near the top of a page, honest structure, and structured data that tells a machine what your page is about.

Tracking is worse than it was. Privacy changes, cookie restrictions and iOS have made attribution genuinely less precise. Anyone showing you a perfectly clean multi-touch attribution model is showing you a model, not a measurement. The workable response is to watch a small number of real business outcomes, enquiries, qualified leads, closed revenue, rather than chasing perfect channel-level credit.

Production is cheap; judgement is not. Anyone can generate a hundred articles this afternoon. That has made volume worthless, and made specificity, real numbers, real constraints, a real point of view, the only thing that still earns attention.

What to have in place before you spend anything

Three things. Without them you are not marketing, you are buying traffic.

1. Conversion tracking that works. You must be able to see which activity produced an enquiry. Not sessions. Not impressions. Enquiries. Without this, every optimisation is guesswork wearing a dashboard.

2. One owner. One person accountable for the outcome, internal or external. Marketing split across three people who each do a bit of it is the most reliable way to spend money and learn nothing.

3. One number that defines success. Cost per qualified lead, cost per acquisition, or return on ad spend: pick one, agree it in advance, and judge everything against it. A programme with five equally important metrics has none.

What it costs in India

Two separate numbers, and any agency that blurs them is worth a second look.

Agency fees. For a small business, a serious monthly programme generally starts around ₹25,000–₹75,000. Mid-market and funded businesses facing real competition typically sit in the ₹80,000–₹1,50,000 range. Enterprise programmes run above that. Below roughly ₹25,000 a month you are buying part of one junior person’s week, a perfectly reasonable purchase, as long as nobody pretends otherwise.

Media budget. Whatever you put into the ad platforms. This should be separated from fees and, wherever possible, paid by you directly from your own accounts. Percentage-of-spend fee models are not automatically wrong, but they do mean your agency earns more when you spend more, and that is worth naming out loud before you sign.

How to tell whether it is working

At three months, ask four questions:

1. Is cost per qualified lead going down, flat, or up?
2. Is the mix of where leads come from becoming more diverse, or more fragile?
3. Can anyone name what was learned, specifically, what was tested and what the result was?
4. Does the report arrive on the same day every week without you asking?

The fourth is not a vanity check. Reporting that only appears when chased is the clearest available signal that the work is reactive rather than run.

Where most programmes go wrong

They start with the channel. Someone decides the business needs Instagram, or needs to rank, or needs to be running ads, and the work begins there, before anyone has established which stage is broken, who owns the outcome, or how success will be recognised.

Then traffic arrives at a page that does not convert, the numbers disappoint, and the conclusion drawn is that the channel does not work for this business. Usually the channel was fine. The destination was not.

Start at the destination, work backwards to the channel, and measure one number honestly. That is most of it.

The best digital marketing decision most businesses make in their first year is deciding what not to do.