5 lessons from delivering bad SEO news to executives

Traditional SEO metrics haven’t been good. We don’t need more studies to see what’s happening, but the data confirms it.

Organic traffic is declining for most SEO clients right now. Seer Interactive found that organic CTR dropped 61% for queries with AI Overviews. Executives are watching their dashboards trend downward, often for months at a time.

Most consultants I talk to aren’t prepared for the conversations that come with it. I’m not talking about the diagnostic part. Most of us can figure out why traffic dropped. I mean, the part where you sit across from a CMO and have to explain what’s happening, why, and what you think the company should do about it. That’s a different skill entirely, and we don’t talk enough about it.

I’ve been in SEO for 13 years and have spent the last six running an agency where I personally lead client strategy and present results to senior executives at B2B SaaS companies. The following are the five things I’ve learned about delivering bad news in what is probably the hardest era in which to be an SEO consultant.

1. Executives are more predictable than you think

A few years ago, one of my B2B SaaS clients came to me with a concern I wasn’t expecting. They had gone into their analytics and looked specifically at the performance of our team’s work, separated from the rest of the site’s organic traffic.

The overall numbers we had been reporting looked fine. But when you isolated the work we were responsible for, the performance was flat. It had not grown at all since we started eight months prior.

I looked over the numbers myself, and the client was right.

When I dug into what had happened internally, the picture got worse. My team knew. They had seen that the work was underperforming, but they had made a decision many consultants make: they reported the numbers that looked good and avoided the ones that didn’t. Instead, they kept presenting overall traffic trends without flagging that our work specifically was not delivering.

Nobody wants to walk into a meeting and say, “This didn’t work.” But hiding a failure is often worse than the failure itself. There are two reasons for this.

• The client will eventually find out. Mine did. And when they did, the damage to their trust in us was not about the underperformance. It was about the fact that we had either not caught it, or, worse, had not surfaced it for them.
• When you hide what isn’t working, you lose the opportunity to show the thing executives actually value most: that you’re able to recognize a problem, diagnose why it happened, and bring a revised plan to the table.

This experience changed how I run every client engagement. I started by rebuilding our reporting to isolate the performance of our own work from overall site trends, and then I implemented the rule that underperformance gets surfaced early, with a diagnosis attached.

Anecdotally, every executive I sit across from has been burned at least once by vendors who obscured their bad results. It’s the consultant who surfaces problems early and brings a plan to fix it who is doing something genuinely rare.

Executives who reacted the worst to bad news were never the ones who received it directly. Instead, they were the ones who were left to discover it themselves, or who could tell that I was dancing around something and trying not to state the bad news directly.

2. Diagnose before you communicate

Early last year, a prospect came to me concerned about a traffic decline. Their internal team assumed AI Overviews were eating their clicks. It makes sense, since that seems to be the default explanation for every problem right now, and sometimes it’s correct.

But I’ve learned not to walk into a room with an assumption when I can walk in with a diagnosis. Before I said anything to the client, I had a look at their site myself. The first thing I wanted to see was whether there were actual keyword losses, and, if so, who replaced them. 

• If competitors had taken those positions, then that would be an SEO problem. 
• If, instead, AI Overviews had absorbed the clicks while rankings held or improved, that would indicate a structural market shift. 

These are two completely different situations that require completely different responses. I wanted to know which we were dealing with so that I could offer solutions from the get-go.

What I actually found was a third issue entirely. The client had run a major PR campaign over the summer that had created a huge traffic spike. The quarter-over-quarter comparison was measuring against that spike, making normal performance look like a decline.

When I pulled the timeline back further and compared pre-campaign to current, the trajectory was actually growth. It was just more stable growth than the spike had made everyone expect.

That diagnosis changed the entire conversation. Instead of walking in and explaining a traffic loss, I walked in and explained what the data actually showed. The client went from concerned to confident in about five minutes. But I wouldn’t have gotten there if I had accepted the surface-level read.

Other times, the diagnosis really is bad news. For example, I had another client with a genuine traffic problem, and, when I dug into it, I found that the cause was technical: a set of pages generating crawl waste that was dragging down the rest of the site’s performance.

Luckily, I had seen this pattern before with another client and knew what the fix looked like. Because of that, I was able to come to the call and say: 

• “I’ve identified what is causing this issue. I’ve seen this pattern before with another client, and I’m going to tell you what we did to fix it, what the recovery looked like, and my theory on what it will do for your site based on the data I am seeing.” 

That ability to diagnose builds trust from the onset of the project.

No executive needs to hear an explanation of crawl budgets or parameterized pages. They need to hear that you’ve found the problem, that you’ve seen it before, and that you have a plan. But you can’t say any of that unless you’ve done the deep diagnostic work first.

The thing that builds confidence in the room isn’t your delivery. Instead, it’s the quality of the diagnosis and the specificity of the plan behind it.

---

3. Surprise bad news and failed experiments are different conversations

There are two types of bad news you can end up delivering to a client, and the conversation you will end up having with your client depends entirely on which one you’re dealing with.

Surprises

The first type is what I think of as surprise bad news. These kinds of problems often happen when the work has been running without a clear strategic structure. You’ve been doing the SEO work. You’re finding opportunities, publishing content, optimizing pages, and staying busy.

But there has been something missing. There hasn’t been a defined plan with specific bets tied to specific outcomes. When traffic starts dipping and someone asks you what happened, you’re left in a difficult position. Why? Because you don’t have a clean way to diagnose what has happened. You weren’t testing a specific hypothesis. You were just doing work.

This happens way more frequently than most consultants would like to admit, especially when there’s no structured review cycle in place. When you’re constantly in “find more opportunities” mode, everything feels productive until the numbers go in the wrong direction.

Failed experiments

The second type of bad news you might have to deliver is regarding a failed experiment. This type is actually much easier to handle than surprises, even though the news might be just as bad.

A failed experiment means you had a plan. You told your clients something like:

• “We are going to try this specific approach because we believe it will produce this specific result.” 

And then it didn’t. But, because you planned it deliberately, that means that you can evaluate the outcome. 

You can tell your client: 

• “These elements performed, these didn’t, here is what the data shows, and here is what I want to try next based on what we’ve learned.”

Taking deliberate bets to weed out surprises

Almost every SEO consultant right now has to report traffic declines to their clients. That’s the reality of the market we’re in. Seer Interactive’s study found that AI Overviews now correlate with a 61% reduction in click-through rates for organic search. 

Yet, executives are seeing these numbers drop and still want to see growth. You’re going to have to deliver that bad news regardless. The difference is whether:

• You’ve already been tracking the trend, forming hypotheses about what’s driving it, and testing responses.
• Or it catches you off guard because you’ve only been watching a monthly report without looking deeply into the underlying patterns.

The best protection against delivering bad news poorly isn’t better communication skills. It’s working in structured cycles where every major effort is a deliberate bet with a defined expected outcome. 

When something doesn’t work, you’re not delivering surprise bad news. You’re reporting on an experiment that all parties were comfortable trying. That’s a conversation most executives are completely comfortable having. After all, they run their own teams the same way.

Once I shifted to this way of working with clients, I saw the dynamic in difficult conversations change fundamentally. But I also learned that, even when you have the right diagnosis and the right framing, when the executive hears the bad news matters almost as much as how they hear it.

4. Never arrive without a recommendation

I’ve sat in enough client meetings to know the exact moment a conversation turns. It’s the moment after the bad news lands, and the client says, “OK, so what do we do now?”

If there’s no answer ready, the room palpably changes. The bad news, which might have been completely manageable thirty seconds before, suddenly feels much worse.

The worst case is when the client finds the problem before anyone on your team does. I’ve seen it happen. The client goes into their analytics, spots a decline, brings it up, and now you, the consultant, are on your back foot.

No diagnosis. No theory about what caused it. Definitely no recommendation. Just scrambling to catch up on something that should have been caught first.

The diagnosis and the recommendation aren’t separate steps. They’re one thing. If you’ve done a deep enough diagnosis to actually understand what happened, you almost always have a theory about what to do next. Showing up without a recommendation means you had not done thorough enough diagnostic work.

These days, before I get on any call where I am presenting something, I make sure I’ve thought through at least two paths forward. Not vague ideas. Concrete options with tradeoffs. I recommend one, explain why, and present the other as a real alternative. The client ends up choosing between solutions instead of sitting with a problem.

For example, I had a client whose legal team was blocking us from publishing comparison listicles, and the whole content strategy — that they had approved — had stalled. Instead of getting on a call with them and saying, “We have a blocker,” I came in with two alternative approaches.

One approach was increasing outreach and placements on third-party listicles. The other was shifting the format entirely, and, instead, focusing on content like analyzing third-party reports in the client’s field instead of creating our own comparison content. The client picked one, and we moved forward. The blocker barely registered because it was already paired with a way through it.

5. The tough conversation builds the relationship

My strongest client relationships were almost never the ones where everything went smoothly. They were the ones where something went wrong, I handled it well, and the client came out of it with more confidence in me than before.

Many of the people these executives work with are spinning results or avoiding hard topics. When someone shows up and says, “This didn’t work, here is why, and here is what I think we should do instead,” it stands out. There’s a difference between transparency that feels like data-dumping and transparency that feels like strategic intelligence. The first erodes trust. The second builds it. 

I used to dread difficult months. Every hard conversation felt like a performance review of my work. After years of watching what actually happens when things go wrong, I now process those moments differently.

Why? Because a smooth month doesn’t illustrate to the client anything about how I operate under pressure. A hard month where I catch a problem, diagnose it, and come up with a plan tells them everything. Those are trust deposits, and they compound over time.

The conversation is part of the work

SEO is getting harder, and, in many cases, the numbers are going to go in the wrong direction.

More of the work is happening in the conversation that follows. Explaining what happened, why, and what to do next isn’t a side skill anymore. It’s a core part of the job.

That means showing up with a clear diagnosis, a point of view, and a plan. It means surfacing issues early, instead of waiting for someone else to find them. And it means treating every dip not just as a performance problem, but as a moment to demonstrate how you operate.

Because, increasingly, that’s what clients are evaluating. Not just the results, but how you handle them.

View the full article