How 202 Digital Reputation built a scalable reputation practice on top of LLM Pulse

Who should read this: reputation, PR and comms agencies who want to run online reputation work in AI answers as a repeatable service instead of a series of one-off audits.

TL;DR

202 Digital Reputation is one of LLM Pulse's earliest and most active partners. They have built an online reputation management practice on top of the platform: they use it to identify the negative sources AI models cite, build cleanup plans for their brand clients, and turn that work into a repeatable service. Their day-to-day feedback has also shaped features that now ship to every LLM Pulse customer, editable sentiment being the most visible one.

About 202 Digital Reputation

202 Digital Reputation is a specialized agency focused exclusively on online reputation management: monitoring, defending and cleaning up how a brand or a person is portrayed across the open web.

AI search widened that job. Managing what shows up in Google is no longer enough. Agencies now have to manage what ChatGPT, Perplexity and Gemini say about a client when someone asks.

The problem: reputation work without a scoreboard

Reputation engagements have always had the same weak point. A team can see that an answer about their client is wrong or unfair, but proving which sources cause it, deciding what to fix first, and showing a client that the cleanup worked are all separate problems.

In AI search that gets harder. The answer changes by model, the citation behind it is often invisible to the client, and a single outdated page can keep resurfacing months after everyone has forgotten it exists.

How they use LLM Pulse

Step 1: Identify the sources

The first step in any engagement is finding what AI models cite about a brand and why the answer comes out the way it does. 202 leans on three parts of the product for that:

• The Citations dashboard, which lists every URL an AI model has cited about the brand, ranked by frequency and weighted impact.
• Sentiment Tracking, for the polarity and topics behind each mention, including which sources drive the negativity.
• Models Comparison, to catch model-specific divergences. Perplexity can be fine while ChatGPT keeps surfacing a 2018 incident report.

Within the first sprint of an engagement, 202 has a list of the sources doing the damage and a triaged plan to address them.

Step 2: Build the cleanup plan

Reputation cleanup is a portfolio of tactics, not one move. 202 updates cited sources where they can influence the publisher, replaces outdated content with fresher and higher-authority alternatives the models will prefer, runs PR and outreach to displace negative citations, and produces on-brand content so the right narratives are easy for an LLM to pick up.

LLM Pulse is the scoreboard for all of it. Every action is measured against the prompts that matter to the brand, week over week, model by model.

Step 3: Correct the classifications that matter

Sentiment classification is hard. Generic NLP gets it right most of the time, but a reputation engagement needs precision, because a single misclassified mention can send an action plan in the wrong direction.

202's daily workflow kept surfacing cases where an automated label was slightly off, and they pushed for a way to correct it without losing the trail. That became editable sentiment: analysts override the classification, the edit stays visibly distinct from the original signal, and both versions are kept for audit.

That feature now ships to every LLM Pulse customer.

Step 4: Feed what they learn back into the product

Editable sentiment is the most visible thing 202 has shaped, but not the only one. Using the product daily against high-stakes reputation work has informed sentiment topic granularity and editing, citations grouping and source-level controls, the sources view used for forensic-style audits, and the reporting flows their brand-side stakeholders read.

Why the model works for them

• Specialization. Reputation is their entire focus rather than one service among many.
• AI-native from the start. Many reputation agencies are still adapting to AI search. 202 has been operating in it since the beginning.
• One playbook per account. With LLM Pulse as the backbone, they can take on more accounts without losing depth on each one: same playbook, same dashboard, same definition of done.

They have grown their LLM Pulse account steadily and now manage reputation across a long list of brands in several countries.

Want to work with 202?

If your brand needs serious online reputation work, particularly around how AI search portrays you, 202 Digital Reputation is a good place to start.

And if you would rather run reputation work in-house on the same setup that powers their engagements, start a free trial or talk to us.