Every owner I meet wants more leads. Almost none of them can tell me what happened to the leads they already paid for.
Open your CRM right now, or your Excel sheet, or the WhatsApp chats your sales fellow keeps on his personal phone, which is where most Indian businesses actually keep their pipeline. Count the people in there who enquired, got a quote, and then went quiet. In every business I have looked inside, from a Ludhiana components manufacturer to a Gurgaon interiors firm, that number is larger than the number of active deals. Usually much larger.
Here is the uncomfortable arithmetic. Every name in that pile cost you money once. You paid for the ad, the exhibition stall, the Justdial listing, the referral dinner. The person raised their hand, which is the single hardest thing to get a stranger to do. And then, because nobody followed up past the second call, they were quietly reclassified from "lead" to "old data" and you went back to the market to buy a fresh stranger.
Reviving someone who already enquired costs you a message and a phone call. Acquiring a new stranger costs you whatever your ad platform is charging this quarter, and that price only moves in one direction. You do not need my numbers for this; pull your own ad spend and divide it by qualified enquiries. Then compare it with the cost of a WhatsApp template message. The old list wins before you finish the calculation.
But, and this is the part everyone skips, you cannot just hand the old list to your sales team and say "call everyone". That is how revival campaigns die in a week. The list has to be sorted first, and the sorting is the actual work.
Why "call the old list" always fails
I have watched this play out. The owner has the same realisation you just had, forwards the Excel to the sales head, and says get on it. Three days later the team has stopped, for reasons that are entirely predictable.
The list is full of numbers that don't connect, people who bought from a competitor two years ago, one fellow who has died, and three who get angry because this is the fourth company calling them this month. The team makes forty calls, gets two conversations and one abuse, and concludes the old list is rubbish. The owner concludes the team is lazy. Both go back to buying new leads, now with proof that revival "doesn't work here".
The list was never rubbish. It was unsorted. Dead and dormant were mixed together, and the dead ones poisoned the effort before the dormant ones could pay for it.
Dead means the reason to buy has expired. Dormant means the reason still exists but the conversation lapsed. The entire game is separating the two before a human being dials a single number.
The sort: four questions before anyone calls
You sort the list by asking four questions of every record, in order. The first three can be answered from data you already have. Only the fourth needs a human.
One: can this reason to buy recur? A CA firm's list of people who enquired about ITR filing is not dead in December; it is dormant until next June, because ITR season comes back every year like the monsoon. The same is true of GST registration enquiries, insurance renewals, annual maintenance contracts, school admission season, wedding-season orders. If the need is cyclical, the lead cannot be dead, only sleeping until the cycle turns. Tag these first and set the date they wake up. This is the highest-value tag in the whole exercise and it requires nothing but knowing your own business.
If the need was one-time and has clearly been met elsewhere, the buyer of a flat who enquired about interiors three years ago and has since moved in, mark it dead and stop grieving.
Two: does the person still exist in that role? For business buyers, the enquiry came from a person at a company, and Indian companies churn purchasing staff constantly. A two-minute LinkedIn check tells you whether the purchase manager who asked for your quote is still there. If she has moved on, the record is not dead, but it is not dormant either; it is a fresh cold lead to whoever replaced her, and it should be worked like one. For consumer records, a WhatsApp check tells you whether the number is even live, which brings us to the next question.
Three: does the channel still reach them? Before any calling, send the whole shortlist one honest WhatsApp template message. Not a pitch. Something like: "You had enquired with us about distributor pricing a while back. We're cleaning our records. Still relevant, or should we close the file?" Watch the ticks. Numbers that bounce are dead. Single grey tick for a fortnight, park it. Delivered and read, even with no reply, means the channel is alive and the record earns a call. Anyone who replies "closed" has done you the favour of sorting themselves; thank them and mark it dead. Anyone who replies with a question just moved to the top of tomorrow's call list.
This one message does more sorting than a week of dialling, because it filters on the cheapest signal available before spending the most expensive resource you have, which is a competent salesperson's morning.
Four: what does the first call say? Only now does a human call, and only the records that survived the first three questions. The call is not "just following up", which is the phrase that has trained all of India to hang up. The call references the original enquiry specifically: the model number they asked about, the quantity, the site location. That detail is sitting in your old WhatsApp threads and quotation PDFs. Using it is what tells the person this is a company that remembers them, not a call centre with a purchased list.
What to automate, and what not to
The tagging, the recurrence dates, the WhatsApp sweep, the sorting of replies into buckets: automate all of it. This is exactly the drudgery software is for, and it is why the sort never happens when it is left to humans with a spreadsheet.
The reactivation conversation itself: do not automate it. I will say that plainly because every tool vendor is selling you the opposite. There are platforms that will happily run an "AI re-engagement sequence" over your whole database, and what they actually do is send five templated messages to people including the dead ones, which converts your one remaining asset, the memory of a decent past interaction, into an unsubscribe. A dormant lead came back to life because a human said something specific to them. A robot saying "We miss you" has never revived anyone, and you only get one chance to knock on that door before it stops being opened.
Automate the sorting. Keep the conversation human. That division of labour is the whole system.
One thing to do this week
Do not buy software this week. Export your list, every source, including the sales team's personal WhatsApp threads, into one sheet, and run question one on it: tag every record where the reason to buy recurs, and write down the month it recurs. That single column, built in an afternoon, will show you more pipeline than your next month of ad spend, and it will tell you whether the rest of this system is worth building.
If the column comes back nearly empty, your business genuinely runs on one-time buyers and your money does belong in new acquisition. At least now you will know, instead of guessing in both directions and paying for it twice.
I write about automation and the systems I actually run at architmittal.com. Originally published there.
