People enquire when they are thinking about the problem, which is rarely between nine and six. They compare in the same session. Whoever answers first is not necessarily the best option, they are simply the one who was there, and being there is a solvable problem that most companies solve by asking somebody to check their phone in the evening, which works until it does not.
Why waiting costs more than it looks
The damage is not the one lead you lose. It is that the lead who waited until Thursday arrives at the call already having heard someone else's pitch, so you are now the second opinion rather than the first. That changes the conversation from 'is this right for me' to 'is this better than the thing I already half-decided'. The second conversation is much harder to win and it is the one slow follow-up puts you in.
Three ways to cover the gap
- A chatbot that actually knows your business. It answers the ordinary questions, captures what matters and books if the person is ready. Cheapest to run, and it fails gracefully: worst case it takes a message. Typical build $2,500 to $6,000 depending on how much it has to know.
- A voice agent that calls the moment a lead arrives. Strongest option when the buying decision is emotional or urgent, because a call at eleven at night when someone has just filled in a form is startlingly effective. Higher build cost and it needs a defined path to a human. Typical build $4,000 to $9,000.
- An automated first reply that is genuinely useful. Not an autoresponder. A drafted answer to what they actually asked, sent in seconds, that buys you until morning. Cheapest of the three and the one most companies should do first.
How to implement it without it going wrong
Start with the message the automation sends when it is unsure, because that is the one that protects your reputation. Decide the handover rule before you build anything: what has to be true for a human to take over, and how fast. Log every conversation and read them for the first fortnight, all of them, because that is where you find the four questions you did not know people were asking. And tell people they are talking to an automated assistant. The version that pretends to be a person is the version that ends up screenshotted.
The arithmetic, which you can do yourself
Take the enquiries you got last month and the share that arrived outside working hours; your form timestamps have this. Multiply by the rate at which you close the ones you reply to quickly, and again by your average deal value. That is the ceiling on what after-hours coverage is worth to you. It will not all be recoverable, so halve it and compare that to the build cost above. If the halved number still clears the build inside a quarter, the decision is not close. If it does not, you have just saved yourself a purchase, which is also a result.