A customer complained online. Who in the company will find out?

When a person calls a company, it is usually clear who should pick up the phone. But what if they wrote a comment under someone else's post, left a review on a map or complained in a local community? This is where the options begin. SMM didn't see it, PR isn't aware, sales doesn't deal with such questions. Meanwhile, the customer is waiting for an answer.

At a large development company we built a system for handling such messages. On average, about 12,800 brand mentions appeared online every month. Going around the internet manually with that volume can take a long time. Especially if you only look where the company itself publishes something.

Here is how this work was organised. The figures below refer to the period described, and the callback results refer to 2025.

"If something happens, someone will send it to me"

The idea of systematic monitoring meets several common objections: there have been no crises so far, acquaintances tell you about important events, and it is unclear how and why to create a separate team.

It is pleasant to argue with the absence of crises. But it says nothing about what will happen next weekend. And acquaintances, however well-meaning, are not obliged to search for mentions of your company around the clock.

Besides, far from every message immediately turns into a big story. A person may write about a broken lift, ask a question about a project or thank an employee. All of this also needs to be noticed. Otherwise management will only hear about the loudest cases, and customers' everyday experience will stay out of sight.

The company introduced a monitoring system in 2018 as part of its reputation work. The task was very practical: find mentions, make sense of them and pass questions on to those who can solve them.

Whether you call it customer service 2.0 or something else is, in my view, secondary. The person wrote where it was convenient for them. You need to organise the work so that their message reaches the company.

What we tracked

Most mentions came from social networks. Analytics helped us understand on which platforms people discussed the company and where official representatives needed to be present.

The share of positive reviews in the monitoring data was 14%. On its own this figure doesn't give the full picture: to draw conclusions, you need to understand the mix of messages and the rules for assessing sentiment.

We were interested in feedback that could be used in our work: where a customer didn't get an answer, which questions kept coming up, what needed improving in the product or service. Reputational threats also required attention, but reducing all monitoring to hunting for scandals would be odd.

Why we needed several tools

We used three types of services (I won't advertise their names). Each had its own task.

The monitoring system collected mentions in social networks and the media. In it you could sort messages by topic, sentiment and platform. This let us see discussions beyond the company's own communities.

Another service collected comments from the brand's communities in a single interface. Specialists could reply to users on different platforms, and the manager could see how quickly a particular employee responded. When comments are scattered across several social networks, a single workspace makes organising the process much easier.

The third tool helped keep data on maps up to date: addresses, phone numbers, opening hours, photos of properties.

Maps were a separate story. Sometimes brokers registered listings for the company's properties in their own name and put their own phone numbers. A person was looking for the developer but ended up calling an intermediary. So checking listings was about both reputation and a very concrete question: who gets the potential buyer's enquiry.

For my own company, I would start choosing services from these tasks. Where do mentions need to be found? Where should employees reply from? Who keeps an eye on property listings? After that, it is easier to assess tools and their cost.

Who was responsible for what

In our company, reputation management sat within marketing and included ORM and SERM. ORM, Online Reputation Management, means managing reputation on the internet. SERM, Search Engine Reputation Management, refers to reputation in search results.

In another company this function may report to PR. Its place in the structure does not solve the problem by itself. It is much more important to agree on who receives a message and where it goes next.

Reputation specialists tracked mentions, made an initial assessment and passed questions on to the relevant departments. There were three main ones: the property management company, the project manager and PR.

The relevant department prepared information or a comment and returned it to the reputation team. The team then continued the conversation with the user. In high-profile and sensitive situations, PR prepared the official position.

This setup is also needed because a specialist replying on social media cannot find out all the circumstances at a property on their own. They need information from the responsible colleagues. Without it, you can quickly write "we have passed on your request" and then spend a long time figuring out to whom.

What deadlines we set

For online communications we had three targets:

• respond to 100% of messages;

• give an initial reply within an hour;

• provide a detailed answer that resolves the issue within 24 hours.

These were our standards. I would not present them as a universal standard for all companies.

The initial reply and resolving the issue are worth tracking separately. Confirming that a complaint has been seen can be done quickly. A substantive answer often requires work from another department (PR, construction, service and so on). If you measure only the speed of the first comment, it is easy to end up with excellent reports and a person who is still waiting for help.

Customer expectations, meanwhile, are shaped outside your industry. If another service answers them within ten minutes, the explanation "that's not how developers do it" is unlikely to please them. So deadlines should be set based on what the team can do and how quickly it gets information from colleagues.

What we found in phone enquiries

Work with enquiries was not limited to social media. The phone remained an important channel, and in 2025 the share of enquiries that got a callback within five minutes grew from 28% to 42%.

When analysing enquiries, we found that 35% came in outside working hours, from 21:00 to 9:00. We set up a night team to handle them.

This example is useful because it shows that the problem can lie in how the work is organised. If a noticeable part of the audience leaves enquiries after the shift ends, simply demanding "call back faster" is not enough. You need to decide who will handle these enquiries.

Where I would start in another company

First, I would check which messages the team sees now. The company's own social media, reviews on maps, media publications and discussions on third-party platforms may reach different employees. It is worth putting this picture together before buying yet another service.

Then I would assign owners and describe how questions are handed over. Who deals with building maintenance, who is responsible for a specific project, when PR steps in, who gets back to the customer with the final answer. Deadlines will also be needed for passing information internally.

Separately, I would check the map listings and how enquiries are distributed by time of day. A wrong phone number and night-time enquiries don't always come up in discussions about reputation, although they directly affect communication with buyers.

And only after that would I choose tools, set standards and go through recurring topics with the responsible departments. Otherwise you can build an excellent system for collecting complaints that nobody uses to improve anything.

In our case, the system linked monitoring to the work of specific services. This principle can be applied in Uzbekistan and other markets too, adapting the platforms, languages and schedule to your audience. You can start by checking a few recent messages: who noticed them, where they were passed on and whether the person got a substantive answer.