Nearby Now
vs RealWork Labs

Whatever they actually did, in as few words as they like. The AI enhancement takes “replaced cap. 3 ton unit, upstairs” and writes it out as prose a homeowner would read, varying persona and length between check-ins on purpose so a city page does not read as forty runs of one template.

Check-in and post text is also checked before publishing for keyword stuffing, repeated phrasing across check-ins, and city-name stuffing. It warns; it does not block. The technician stays in charge of their own words.

An auto check-in can fire when the van enters a geofence around the job, bounded by a maximum per day, a minimum interval between them, and a minimum distance from the last one. The guards exist because uncapped automatic content is its own quality problem.

There is also a reminder that nudges the technician when they arrive, for teams that would rather keep it manual.

Faces and licence plates are detected and blurred automatically. A job photo taken in a driveway routinely catches a homeowner, a child, or a plate, and publishing that to a public page is a problem a contractor should not have to remember to avoid.

Objects in the image are also detected and labelled, which is what lets photo content carry more than a filename.

The destination is decided before the customer rates. Requests go by SMS, email, QR code, and a phone number that records an audio testimonial. A QR review is not counted until the customer verifies by email or SMS, because a yard sign can be scanned by anyone.

Businesses with more than one Google Business Profile have reviewers routed to the profile nearest them automatically. The full decision tree is published.

AI Visibility Tracking, a paid add-on to the account. A fixed set of questions — the ones a customer would actually type, in the cities that business actually works in — put to six answer engines on a weekly schedule: Google AI Overviews, Perplexity, ChatGPT, Gemini, Claude and Grok. Through each engine's API rather than a logged-in session, because a logged-in session carries your history and answers you rather than a homeowner in the city you are asking about.

Every reply is kept, and four things are read out of it: whether the business was named, where it fell inside that answer, which other businesses were named alongside it, and which model said so on which day. An outage, a refusal or a timeout is recorded as “could not check” and kept out of the denominator, because somebody else's bad afternoon is not your client's bad month.

And it prints no score, no grade and no ranking — not on a screen, not in an export, not in an email. What it prints instead is a citation rate, always beside the two counts it was calculated from, and the sampling caveat in full next to the figures rather than in a footnote at the end. The competitor names in each answer are kept too, which is the part agencies end up using most.

This is the sentence the product prints beside every figure — on the dashboard, in the export and inside the monthly scorecard, unabbreviated in all three:

Each figure is a sample: one question put once to one model through its API on the date shown. Answer engines reply differently to different people, in different sessions and as their models change, so this is evidence of what an engine said when it was asked — not a position, a ranking or a score.

$155 to $239 per location per month, on the pricing page, unlimited users, no setup fee, 14-day trial.

Agencies get the deepest white-label layer in the category: branded domains, themes, knowledge bases, and four white-labelled mobile apps live in the app stores under partners' own names.