| Questions to think about as you read As you read, think about a change your team is considering — or one you’ve made recently. What problem were you trying to solve? What was the current approach doing well that you didn’t want to lose? And how confident were you that the replacement would actually produce a better outcome, rather than simply being newer, cleaner or more appealing? You might also ask yourself: what are you measuring today, and what might be valuable but harder to measure? |
I love the Kennedy Center. I’ve lived within walking distance of it for more than 20 years, and during that time it’s become part of my life in Washington, DC. I’ve gone there with family and friends for plays, musicals, concerts, comedies and tragedies; sometimes I’ve walked over without tickets to anything, just to wander around and enjoy the building and the views.
So I’ve been following what’s happened there lately with more than casual interest. It’s sad for me to watch. I’m not going to get into the politics of it here; that’s not what this article is about. But watching all the changes got my marketer brain going.
The Kennedy Center wasn’t perfect before. Its aging building had documented maintenance and infrastructure needs; its FY2025 budget request included increased maintenance on aging infrastructure and work to replace and modernize building systems. There were things that needed attention and things that could be improved.
Then people set out to change things. And things changed. A lot.
That got me thinking about something I see in marketing all the time:
We are very good at identifying everything that’s wrong with what we have.
We’re often much less rigorous about identifying everything it’s doing right before we change it.
“This isn’t working” isn’t a strategy
You’ve probably been in one of these meetings. “Our website is dated. We need a redesign.” Or, “Our email templates are boring. We need something fresh.” Maybe it’s, “This platform is terrible. We need a new one.” Or, one of my favorites, “These subscribers haven’t opened an email in six months. Let’s get rid of them.”
Maybe every one of those statements is true. Maybe the website is dated, the email template does need a refresh, or the platform really is terrible. But “this isn’t working as well as we’d like” and “this change will make it better” are two entirely different propositions. We have a tendency to treat them as if they’re the same. They’re not.
We often compare reality to fantasy
Think about how we evaluate the marketing program we have today. We know all its flaws: the website pages that drive us crazy, the reports our ESP can’t generate, the convoluted workaround Chris in marketing ops has to do every Tuesday morning, the emails that aren’t performing as well as we’d like.
We have years of evidence about everything that’s wrong with our current reality. Then someone shows us the replacement. The new website mockups are gorgeous, the new email template looks fresh and modern, the new platform promises to automate everything Chris has been doing manually, and the new channel is where our audience supposedly spends all its time.
See the problem? The incumbent has a rap sheet. The replacement has a sales deck.
We’re comparing the actual, messy reality of what we have with the idealized future we’re imagining. That’s not a fair fight.
The better comparison is the actual current state — its benefits, costs, limitations and risks — versus the likely future state, with its benefits, costs, limitations and risks. And that second equation is a lot harder to fill in.
Before you fix what’s wrong, inventory what’s right
This may sound obvious, but I don’t think it is. Let’s say you’re redesigning your email program. The current templates are dated, they’re not especially attractive, and maybe the design isn’t consistent with your current brand standards. Time for a redesign, right?
Probably. But before you start, what else do you know about those ugly old templates?
Maybe subscribers recognize them immediately. Maybe the call-to-action buttons perform extraordinarily well. Maybe the hierarchy makes the messages easy to scan. Maybe they render beautifully in the email clients your audience actually uses, and customers have learned exactly where to look for the information they want.
None of those are reasons to keep an outdated template forever. They’re reasons to identify the things that are working before you change the things that aren’t. Otherwise, you can end up with a beautiful new template that meets every internal design objective… and generates less revenue.
Sigh.
Things usually do more than one thing
Here’s another trap: we tend to define something by the problem we’re trying to solve. Those inactive subscribers? They’re “dead weight.” That expensive martech platform? “It’s costing us too much.” Those repetitive emails? “Nobody needs all those.”
But things in a marketing ecosystem rarely perform just one function. An email newsletter may generate clicks, but it may also keep your company top-of-mind with people who won’t need your product for another six months. A subscriber who hasn’t clicked recently may still be reading, purchasing through another channel or forwarding your emails to someone who does.
Your expensive technology platform may contain years of integrations, workflows, data and institutional knowledge that aren’t included on the vendor comparison spreadsheet. Your brand identity may look stale to the people who stare at it every day while being instantly recognizable to customers who see it twice a month.
This doesn’t mean you shouldn’t change any of these things. It means you need to understand everything they’re doing before you decide they’re expendable.
Cutting isn’t necessarily optimizing
This one is especially important for marketers because we love measurable things: cost per acquisition, click-through rate, conversion rate, revenue per email, cost per lead. That’s usually a good thing. I’m a data person. Give me a spreadsheet and I’m happy.
But measurable doesn’t necessarily mean important, and important doesn’t necessarily mean easily measurable.
Suppose you cut $100,000 from your marketing budget. Congratulations! You saved $100,000. Unless that $100,000 was generating $300,000 in incremental profit. Then you didn’t optimize anything; you just cut spending.
Real optimization means improving the relationship between what you put in and what you get out. If you achieve the same or a better outcome with fewer resources, that’s efficiency. If you use fewer resources and get a significantly worse outcome, that’s just fewer resources. There is a difference.
Some changes are harder to undo than others
Changing a subject line strategy? Easy to reverse. Trying a different cadence? Also pretty easy. Removing 400,000 subscribers from your database requires more thought. Replatforming your entire marketing technology stack requires a lot more thought. Rebranding a company? You get the idea.
The harder something will be to undo, the more evidence I’d want before doing it, because optionality has value. If you test a new approach with 10% of your audience, you still have choices. If you roll it out to everyone and dismantle what came before, you have fewer.
This is one of the reasons I’m such an advocate of testing. It lets us learn without having to bet the whole program on our hypothesis.
Seven questions to ask before making a big marketing change
So here’s the checklist I’m taking away from all of this. Before you significantly change something in your marketing, ask:
1. What problem are we actually trying to solve?
Not “we hate the website.” What’s the business problem? Low conversion? Poor navigation? Slow load times? Weak lead quality? Be specific.
2. What evidence tells us this is a problem?
Opinion is useful. Data is better.
3. What is the current approach doing well?
Force yourself to answer this one. What do we absolutely not want to lose?
4. What else is it doing that we may not be measuring?
Think beyond its obvious function. What other channels, people, processes or outcomes depend on it?
5. What are we assuming the replacement will do better?
Write those assumptions down. Seriously. Because six months from now, everyone will remember being much less confident about them than they actually were.
6. How reversible is this decision?
The less reversible it is, the higher the bar should be.
7. How will we know we’re better off?
This may be the most important question. Launching the new website isn’t success. Migrating to the new ESP isn’t success. Implementing the new email template isn’t success. Those are outputs.
What outcome should improve? Conversion? Revenue? Engagement? Customer effort? Retention? Profit? Define it before you start, then measure it afterward.
Curiosity before intervention
None of this is an argument against change. Please don’t leave this article thinking Jeanne said you should keep your terrible website forever. I did not.
Optimization requires change; sometimes substantial change. But good optimization begins with curiosity. Instead of “This isn’t working. Let’s replace it,” try asking, “Why is it working the way it is?”
Then ask what’s working that you don’t want to lose, what you believe will happen if you change it, and what’s the safest way to find out whether you’re right. That’s a very different conversation, and probably a much more productive one.
Because “imperfect” isn’t the same thing as “worse than the available alternative.”
Finding something that’s broken is only half the job. Before you fix it, understand what the current system is doing well, including the things you aren’t measuring. Change isn’t optimization simply because something changed; the new version has to produce a better outcome.
Your marketing almost certainly has things that aren’t working as well as they could. Fix them… but make sure you know what isn’t broken before you do.
| Your turn I’d love to hear how this shows up in your work. Have you ever “fixed” something in your marketing only to discover later that you’d weakened or removed something that was working? Or have you successfully protected something valuable while making a major change? Email me and tell me what happened — what you changed, what you learned, and what you’d do differently next time. I read the replies, and I won’t share your response publicly or quote you in a future issue without your permission. |
Go Deeper (Articles Related to The Big Idea)
Better Marketing Starts After the Results Come In
Don’t stop at “that worked” or “that didn’t.” Ask why. The more you understand what drove the result, the better your next decision will be.
Read the article on Martech.org →
Why Most Email A/B Tests Fail (and How to Fix Them)
A/B testing only helps if the test is designed to teach you something. Start with a real question, define what success looks like, and use the answer.
Read the article →
Marketing Foundations Outlast Marketing Fads
New tools are useful. New tactics can be exciting. But neither replaces the need to understand your audience, your message, and what you’re actually trying to improve.
Read the article →
Also Worth Your Attention
New on MarTech
The problem with AI doing exactly what you ask
AI doesn’t need to “go rogue” to create problems. Sometimes it just needs a poorly defined objective and the ability to pursue it very effectively. My latest MarTech article looks at what that means for marketers working with increasingly capable AI agents.
Read the article on MarTech.org →
New on HubSpot
The Best Free Email Marketing Software: An Honest Roundup
“Free” email marketing software often comes with strings attached — contact limits, send caps, restricted automation, or features that disappear unless you upgrade. I looked at the tradeoffs across several popular tools to help marketers understand what’s actually free, where the limits kick in, and which option may fit their needs best.
Read the article on the HubSpot blog →
Workshop | Virtual | October 5-6, 2026
More Effective A/B Split Testing
Good testing doesn’t just tell you which version won; it helps you understand why it won, so you can apply that learning again. In this hands-on online workshop, I’ll walk through how to develop stronger hypotheses, structure cleaner tests, find worthwhile ideas to test, and build an ongoing testing program that improves performance over time. The next session is October 5–6, 2026.
Save 20% when you register by September 18, 2026 →
Digital Summit | Philadelphia, PA | September 23, 2026
Stop Optimizing Emails. Start Managing Your Email Portfolio
I’ll be at Digital Summit Philadelphia next Wednesday, September 23, talking about how to step back from individual campaigns and evaluate your email program as a portfolio of investments. I’ll share a practical framework for deciding what to keep, optimize, fix, or retire — and how to identify the email opportunities with the greatest potential business impact.
See the session details and register →
ESPC 2026 Fall Meeting | Washington, DC | November 3, 2026
Deliverability Is Changing Quickly — Here’s What Marketers Need to Know
I’ll be joining deliverability experts to unpack what’s changing across Gmail and Yahoo sender requirements, authentication, complaint thresholds, unsubscribe rules, bounce codes, open rates, Postmaster Tools and more. We’ll focus not just on what changed, but on what it actually means for your email program — and what you should do about it.
See the session details and register (it’s free) →
MarTech Conference | Virtual | November 4, 2026
Beyond the Opt-In: Engineering Consent for Better Marketing
I’ll be joining Joss Merckaert and Sal Tripi for a MarTech panel on how marketers can move beyond treating consent as a compliance checkbox and start using it as a strategic capability. We’ll look at how better preference management, transparency and customer choice can improve first-party data, strengthen personalization and AI, and build deeper trust with customers.
Learn more and register (it’s free) →
Until next time,
jj
Jeanne Jennings is the Founder and Chief Strategist at Email Optimization Shop, a boutique consultancy and training organization where she helps clients craft more effective and more profitable email programs.
Learn more at www.EmailOpShop.com and sign up for our free newsletter to get more content like this.
Author’s Note
I believe in being transparent about my use of AI. I used ChatGPT to help organize my thoughts, conduct background research, and develop an early draft of this article. The opinions, perspective, and final editorial decisions are mine.
