Recently, I was part of a conversation about a client experience and let’s just say it had not gone the way we wanted.
As we were diving in to understand what happened, someone shared a piece of context that was supposed to explain why the feedback was surprising.
“This was the exact same approach we used last year and the client said they loved it.”
Not just loved it, we were their favorite agency partner ever last year (or something like that) and we received nearly perfect reviews.
At first glance, that sounds like pretty compelling evidence in the team’s favor. We know this works.
Except there was something about this situation that had the opposite effect on me.
The second I heard we were using the same approach as last year, I stopped immediately. What if that is the problem in the first place?
Not because the previous feedback was wrong or new stakeholders or anything like that. I feel pretty confident the deliverable last year was indeed excellent.
But a great score from the past only proves one thing. That it worked in the past.
It does not automatically mean it is the right thing to keep doing. The more I thought about it, the more I realized how often we confuse those two things.
Best practices have a half-life
We love best practices.
It is one of the most reassuring aspects of work. Something works. We learn from it. We document it. We scale it. That is how organizations get better.
But best practices become routines. As teams, we use the same reporting structure year after year even though the metrics and capabilities have completely changed. I get that it takes time to align stakeholders on the KPIs that matter most but let’s make sure those KPIs are the best available to actually correlate to business outcomes.
Whether it is a long-standing meeting that has lost the plot, a pitch deck that got repurposed one too many times or just any of those types of projects (you know what I mean). This is how we do it, this is how everyone likes it, etc. Well, have you proposed a stronger solution?
The conditions for success are constantly changing. Think about the bar for good commerce reporting in 2026.
Retailers are making incredibly rich purchase signals more readily available than ever before. Amazon Marketing Cloud is one of the more powerful and connected clean rooms out there. Walmart has rolled out APIs for Scintilla. And there are countless other opportunities as well from Kroger, Instacart, DoorDash and other developing data solutions.
The tools are getting better while the barrier to using them gets lower. That means the bar for good is climbing through the roof.
Sometimes success is the thing you have to question
Here’s a bit of the uncomfortable part. We all hear so much about learning from failure. And if you are reading this newsletter, you are probably the type of person that uses failure to get curious.
You are also a human being and like all human beings, if we get top marks on something, we are highly likely to preserve that approach. Words of affirmation are my love language and so if someone looks at a report I made and says great job, I am extremely likely to keep doing it the same way to get the same result.
That is completely rational.
But the question I have learned to ask is not, did this work. It is under what conditions did this work. And are those conditions still true today?
A media strategy can be brilliant one year under one market condition and look like you completely missed the mark the next year. Look how much discussion there has been around GEO and Sponsored Prompts in the last year alone.
The past is useful evidence. We just can’t ever let it get veto power over what will drive the business forward.
Re-pitch the client
There is a topic I have written about extensively to internal teams throughout my career: I think we should re-pitch our clients.
Not literally tell them to go give us an RFP but I firmly believe that mentally, at least once a year, we should be asking ourselves as a collective group some really critical questions and see what comes back.
If we had to win this business again today, would we show up the same way? Would these still be our recommendations and priority deliverables? Are we getting value out of our current meeting cadence? Do these insights matter? What are we not doing?
Clients are not paying agencies because they want to keep running the same play again and again and again. There is an AI for that.
They want direction. An inspirational partner to push the thinking and the ideas. And as lofty as that may sound, frankly it’s just a partner with their hands on the wheel, helping them understand what changed and what they should do differently because of it.
The benchmark cannot only be, was this better than last year’s version. I like to challenge teams with the question: What is possible now that wasn’t possible before?
AI raises the bar for the humans
In a world of AI, more is more.
With LLMs and the latest data science capabilities available, we can process insane volumes of data. We can now review significantly larger data sets AND reduce the time we spend monitoring them, spotting anomalies and drafting summaries.
But I think that undersells what is happening. That really only makes us more efficient at capturing today’s value.
If technology can watch, detect, summarize and report, then the human value gets to move somewhere else. Judgment and discernment.
Time gets reinvested to pressure test the context. Connect things that we are seeing with our own eyes and human experience. Know when the signal feels misleading, understand the consumer truths and make a hard recommendation when the obvious choice is staring right in front of us.
That time (and mental energy) can go toward something new. Getting curious, discovering new metrics and unearthing new insights to drive the business forward.
You have to let go to grow
Forget about business for a second. I care about this deeply on a personal level. Doing the same thing over and over again sounds like an incredibly boring way to live life.
I may be a little crazy but I just absolutely love the craft of work (I mean I did start a newsletter in my free time to write about work).
There is such a mentally stimulating aspect of innovation. What did we learn? What changed? What can we try? What has technology enabled now that we could not do before?
I don’t want to show up every morning and run the exact same ten plays because they worked last season. I want to know what the next play should be.
It can be incredibly hard to let go. Particularly if something worked really well. It’s easy to move on from failure because you have immediate permission.
But how do you move on from success when every other reason points to defending the thing?
Would we build it this way today?
I keep thinking about the comment that we had nearly perfect scores from the client.
I still think it was an important signal. We created value in that moment.
But it was just that, a moment, and certainly not permission to stop evolving.
Here is the question I asked back to the team: If we were starting today, knowing everything we know and using everything that is possible now, would we still build it this way?
If the answer is no, just let it go.