How often to scan groups so you stop losing enquiries

An enquiry has a shelf life

A post where somebody is looking for a contractor is interesting for a few hours. Then it drops below newer entries, a dozen replies sit under it, and its author has already arranged something.

So frequency is not a comfort setting, it is a competitive variable. The same group checked every two hours and checked once a day are two different products as far as your salesperson is concerned.

The group sets the pace, not your preference

An active local group can produce enquiries all day. A trade page produces a few a week. One schedule laid over both means losing enquiries in one and paying to check an empty room in the other.

In practice: spend the first week watching when questions actually appear in each source, and only then tighten the schedule where something is happening.

Which is why the schedule belongs to the source

In LeadScanner every source carries its own schedule and its own frequency. You pay for scans, so moving budget from a quiet page to a lively local group is a decision you make, not one a single account-wide slider makes for you.

The rule works the other way too: a source that produced nothing for a month is worth slowing down or switching off, rather than paying for out of habit.

Why nobody should promise a number of leads

How many enquiries come out of your groups depends on how active they are, how narrowly focused, and whether people there ask for contractors at all. No vendor knows this before the first scan runs — us included.

There is one sensible way to check: take the trial on three of your own sources, count the first twenty matches and see how many you discard. If it is more than half, either the keywords are wrong or those groups are not what they appear to be.

A guaranteed number of leads in this category is a promise made without grounds. Read it as a warning sign, not a selling point.

What to do with a group that is too active

A very lively local group is often the hardest source, because it drowns in its own traffic: alongside requests for contractors come classifieds, memes and arguments about parking. Tightening the schedule alone in a place like that raises the cost without necessarily raising the number of useful matches.

The better order is the other way round. First narrow what you are looking for — the specific service, the specific area, the specific phrasing people actually use when they ask for a recommendation. Only against a filter that narrow does raising the frequency pay, because each additional visit then brings enquiries rather than noise.

The second setting worth switching on straight away is quiet hours. An enquiry written at two in the morning will wait until breakfast anyway, and a notification sent at that hour teaches your salesperson to mute the channel — which costs more than one late lead.

Run a one-week baseline

Use the first week to measure, not optimize. For every match, note the source, posting time, matched phrase, service requested, and whether it was relevant. Add the time when someone first saw it and the time when you replied. Keep the schedule and phrases unchanged during that week. Changing both at once makes the result impossible to interpret. At the end, sort sources into regular relevant requests, occasional relevant requests, and mostly irrelevant matches. This record gives you a starting point for schedule changes and shows where phrase work is needed first.

At the end of the baseline, compare sources using the same simple totals. Count all matches, relevant matches, replies sent, and conversations that moved forward. A source with fewer matches may still be stronger if most requests fit your work. Check the time gap between a post and your first review. If relevant requests repeatedly wait until the next workday, the schedule or notification coverage is too slow. If scans repeatedly find nothing, reduce that source before reducing a useful one. Keep this review in one shared sheet.

Change one variable

Make schedule changes one source at a time. First decide what problem you are solving: late discovery, too much noise, or no relevant demand. For late discovery, increase checks only where relevant requests appear regularly. For noise, change the matched phrases before increasing frequency. For weak demand, slow the source and test another relevant public source. Leave the revised setting in place long enough to gather a comparable set of matches. Then compare it with the baseline. A change that creates more cards but no more relevant conversations is not an improvement. Write down the date and reason for every change.

A practical example is a neighborhood group that produces many general recommendations. Your broad phrase may catch requests for services you do not provide. First add the exact service and local area people use in their posts. Review the next set of matches. If the remaining matches are relevant but arrive after competitors reply, increase that source’s frequency. If the revised phrases return little, do not compensate by checking more often. Move attention to a source where people use clearer buying language.

Set a response handoff

A scan only helps if somebody owns the next step. Decide who reviews new lead cards, who checks whether the request fits, and who writes the reply in the public thread. Give each person a backup for absences and busy periods. The reply draft can speed up review, but it still needs a human check for accuracy, tone, and availability. Set notifications for the hours when that person can act. If nobody is available overnight, hold those notifications until the working day. Review unanswered relevant matches during the daily check-in.

Use a short review checklist before replying. Confirm that the post asks for your service, the location is workable, and the request is still current. Read earlier replies so you do not repeat information already given. Then personalize the draft with one useful detail, such as your service area or the next available step. Do not post a generic sales message under every request. Record the outcome as replied, not a fit, already solved, or no response. Those labels show whether the source, the phrase, or the handling was slow.