Reputation management for small business: how it works and how to set it up

Reputation management for small business comes down to three habits: ask every client for a review, reply to every review that arrives, and catch problems before they become public. When all three run from the same system as your CRM and inbox, your online reputation stops depending on who remembered to do what. This guide covers how to set it up on the levelpartner.ai platform, what it takes to run each week, and where firms usually go wrong.

What good looks like

A prospect who searches for your firm sees a steady stream of recent reviews, each with a reply from someone at your company. Complaints are answered calmly and taken offline. Your team sees every new review the day it lands, alongside the client's full record, so the person replying knows the history.

That is the goal for an established company with 10 to 250 staff. It is less about star ratings and more about showing a prospect, in under a minute of reading, that you look after clients and answer when something goes wrong.

Before you start

Gather these before you set anything up:

• Verified business listings for every location, starting with your Google Business Profile.
• Admin access to the platform's Settings and Reputation sections.
• A defined finish point for a satisfied client in each service line, which will trigger review requests.
• A short reply policy: who replies, how fast, and what tone to use.
• A named owner for complaints, usually an operations lead or office manager.

How to set it up

1. In Reputation, click Connect Review Platforms in the setup wizard.
2. Select Google or Facebook and follow the connection prompts.
3. In Settings > Review Link, confirm and copy your review link.
4. On the Review Requests page, choose a contact.
5. Send the review request by SMS or email.
6. In Settings > Reviews AI, toggle AI responses on.
7. Select your preferred tone and response behaviour.
8. On the Widgets tab, choose a carousel, grid or list widget.
9. Copy the widget code into your website's HTML.

Once the basics work, automate the request with a workflow that fires when an opportunity reaches your satisfied-finish stage, and skip any contact tagged with an open complaint. A second workflow can create a task for the complaint owner when a low rating arrives. Keep AI responses on approval until you trust the tone, and leave complaints for a person to answer.

The three habits in practice

Asking every client

Most firms ask the clients they like and forget the rest. The workflow fixes that by asking every client who reaches the finish point, in the same words, at the same moment. Our step-by-step guide on getting more Google reviews with an automated request covers the request itself in detail, including how to write it.

Replying to every review

Replies are read by prospects more than by the reviewer. A thank-you to a positive review should be personal and short. A reply to a negative review should acknowledge the issue, avoid arguing the detail in public, and give a direct contact to resolve it. AI content on the platform can draft a reply for your team to edit, which saves time, but a person should read and approve every reply before it is posted. Our guide on how to respond to negative reviews sets out a reply structure and who should own replies.

Catching problems early

The best negative review is the one that never gets written. A short satisfaction check sent before the review request, using a form or survey from Sites, gives unhappy clients a private route to tell you. Route those answers to the complaint owner as a task. Our workflow automation examples include patterns for this kind of escalation.

Be careful here: the satisfaction check must not decide who receives the review request. Sending review links only to happy clients is known as review gating, and it breaks Google's review policies. The FTC's endorsement guidance treats it as misleading in the US, and the UK's DMCC Act rules on misleading review information have applied since April 2025. Ask everyone, at the same time whatever their answer, and use the check to fix problems, not to filter.

What this replaces and what it costs in time

Without a system, reputation work is a scatter of habits: a partner who replies to reviews when they notice them, a manager who asks a few clients for a review after a good meeting, and a complaint that surfaces on a review site before anyone inside heard about it.

The platform replaces that with one review inbox, one request workflow and one escalation route. Any follow-up messages from reviewers arrive in the unified inbox with the rest of the client conversation. The weekly routine is short: reading new reviews, approving replies and checking that complaints were closed. Most of the setup effort goes into writing the messages and agreeing who owns replies. If you show reviews on your own site, a Google reviews widget keeps that page current without manual updates.

When not to rely on this alone

Reputation management works on the reviews you earn. It does not fix a service problem. If the same complaint appears three times in a quarter, the answer is an operational change, not a better reply template.

It is also not the right tool for every review site. Some industry directories and regulated comparison sites have their own rules on requests and replies. Check each one before adding it to an automated sequence, and keep requests to sites where you are confident of the rules.

Troubleshooting

New reviews are not showing in the platform

The listing connection has usually expired, or it was made with an account that does not manage the location. Reconnect it in Settings with the right account and check Reputation again the next day.

The complaint task never fires

The rating condition in the escalation workflow is often set the wrong way round, or the trigger is watching the wrong location. Open the workflow, check the condition reads "three stars or fewer", and test it with a sample review record if your setup allows it.

Replies are going out inconsistent in tone

Several people are replying without a shared guide. Write five model replies, for a five-star review, a mixed review, a complaint about price, a complaint about delay and a review from someone who was never a client, and ask everyone to start from those.

Review requests reach clients with an open complaint

The skip condition depends on a tag that nobody is applying. Make the complaint form or task add the tag automatically, so the workflow can see it without anyone remembering.

Frequently asked questions

How much time does reputation management take each week?

Not much once it is set up: a fixed weekly slot to read new reviews, approve replies and check complaints are closed. The time grows with your review volume and the number of locations. The request workflow itself runs without anyone touching it.

Should we reply to every review, including old ones?

Reply to every new review from now on. For older reviews, reply to any unanswered complaint from the last year and leave the rest. A visible pattern of recent replies matters more than catching up on everything.

Can AI write our review replies?

AI content can draft a reply that your team edits and approves. Do not let replies post without a person reading them, because a wrong detail in a public reply is hard to undo.

Where should we start if we have very few reviews?

Start with the request workflow, because volume comes before everything else. Our guide to setting up review requests walks through it, and our guide to sending an automated text message covers the SMS side, including consent.

The journey this guide describes