Solved: Best Way to Switch to Flexible Google Ads?

Hi there,

Thanks for reaching out! Happy to give you some initial thoughts on your question about switching to flexible ads. It's a common question, and the 'best' approach isn't quite as straightforward as Google's reps might make it sound. My gut feeling is that your instinct about losing your existing campaign data is right, and simply duplicating everything is probably not the way to go.

I've put together a pretty detailed breakdown of my thinking on this, how I'd approach it, and some of the wider strategic points you should be considering. It’s a bit of a read, but I reckon it’ll give you a much clearer picture.

We'll need to look at why you shouldn't just assume flexible ads are better...

Alright, first things first. Let's challenge that core assumption: "I'm sure they will perform even better with flexible ads." I hear this a lot, and it's exactly what Google wants you to think. They push Responsive Search Ads (RSAs) hard, and they're even making them the default. But are they magicaly better? Not always. In my experience, it's a real mixed bag.

You have to remember what an RSA is doing. You're giving Google a big box of Lego bricks – up to 15 headlines and 4 descriptions – and you're telling its algorithm, "Here, you figure it out." The system then runs thousands of tiny experiments, combining these assets in different ways to see what works for different users, keywords, and devices. On the surface, this sounds amazing. Automated optimisation, right? The problem is, it’s a bit of a black box.

With your current ads, which I assume are Expanded Text Ads (ETAs), you have total control. You know exactly what Headline 1, Headline 2, and Headline 3 say, and in what order. You've crafted that message perfectly. With an RSA, you lose a lot of that control. Sure, you can 'pin' headlines to certain positions, but the core idea is to let the machine take the wheel. This can lead to a few issues:

• -> The Risk of 'Averaging Out': The algorithm isn't necessarily aiming for the single best-performing combination possible. It's aiming for a stable, predictable, *average* performance across a huge number of impressions. For every brilliant combination it creates, it might also create dozens of mediocre or nonsensical ones. Your handcrafted ETA, which has been running well for months, might already be outperforming this 'machine average'. By switching wholesale, you risk trading a known winner for an unpredictable average.
• -> Loss of Clear Messaging: Sometimes the combinations Google creates just don't make sense grammatically or contextually. A headline that works perfectly in position 1 might sound odd in position 2 next to a different headline. This can weaken your message and make your ad look sloppy, which is the last thing you want.
• -> It Benefits Google As Much As You: Why does Google push RSAs so hard? Because it gives their system more data and more flexibility. An RSA can be eligible to show up in a much wider variety of auctions than a rigid ETA. This means more impressions, more clicks, and ultimately, more ad spend for Google. While that *can* also mean more traffic for you, it doesn't automatically mean more *profitable* traffic.

You've got campaigns that are already performing well. That's a fantastic position to be in. It means you've already found a combination of targeting, bidding, and ad copy that works. The goal now is incremental improvement, not throwing the whole system out and starting again based on an assumption. The biggest mistake I see people make is abandoning something that's working for something new and shiny that promises the world. So, let's not switch. Let's test.

I'd say you should test, not switch...

This brings me to the core of your question: what's the best approach? The answer is simple and absolute: test the new format within your existing campaigns. Do not, under any circumstances, just duplicate the entire campaign.

Your concern about losing the data and optimisation history is spot on. That history is gold. For the past 3-4 months, Google's algorithm has been learning about your campaign. It knows which users are most likely to convert, what times of day are most effective, which devices perform best, and what your Quality Score is for your keywords. This data is tied to the campaign, ad group, and keyword level. If you duplicate the campaign, you are literally throwing all of that learning in the bin and starting from scratch.

The new campaign will enter the 'learning phase' all over again. Performance will be volatile, costs might spike, and it could take weeks to get back to the stable performance you're seeing now. It’s an unnecesary step backwards.

Furthermore, duplicating creates a flawed experiment. If you run Campaign A (old ads) against Campaign B (new ads), you've changed too many variables.
-> Is Campaign B's performance different because of the flexible ads?
-> Or is it because it's a new campaign in the learning phase?
-> Or is it because it's entering auctions at slightly different times and getting slightly different competition?
You'll never truly know. You can't draw a reliable conclusion. The only way to conduct a clean, scientific test is to isolate the variable you want to test. In this case, that variable is the ad format.

The correct method is A/B testing at the ad group level. Here’s why this is the superior approach:

• -> It Leverages Existing History: By adding an RSA into an ad group that's already running, the new ad benefits from the entire performance history of that ad group and its keywords. It's not starting from zero. It has a much better chance of performing well, quickly.
• -> It's a Direct, Fair Comparison: Inside a single ad group, Google will rotate your ads (your old ETAs and your new RSA) against each other in the same auctions, to the same audience. It’s a true apples-to-apples comparison. After a week or two, you'll have clean data showing which ad format delivers a better conversion rate or a lower cost per acquisition. There's no guesswork.
• -> It Minimises Risk: You're not pausing what's already working. You're simply adding a new challenger into the ring. If the RSA is a complete flop, no big deal. Your trusty old ads are still running and carrying the weight. You can just pause the RSA and you've lost very little. If you'd duplicated the whole campaign and paused the old one, a failed test would be far more damaging.

To make this decision process clearer, I've mapped it out in a simple flowchart. It visualises the logic we just walked through and should make it obvious why in-place testing is the only logical path forward.

You probably should structure your tests properly...

Okay, so we've established that we're going to test within your existing ad groups. Now, how do we do that effectively? Just throwing in a bunch of random headlines won't cut it. A well-structured test will give you clearer results, faster.

Here’s the process I'd follow:

1. Choose Your Battlefield: Don't roll this out across every single ad group at once. Start with your top 2-3 highest volume, best-performing ad groups. These are the ones with enough data to give you a statistically significant result quickly. Pick the ones that consistently drive conversions at a good cost.

2. Keep the Champion Running: Do not pause your existing ads in that ad group. Let them run alongside the new RSA. The old ads are your control group, the baseline against which the new ad will be measured.

3. Build Your First Challenger (The RSA): Now, create one new Responsive Search Ad in the ad group. The key here is to fill it with high-quality, relevant assets. Don't just copy-paste your old headlines. Think about variety. You want to give the algorithm good material to work with. Your assets should include:

• -> Keyword-focused headlines: Headlines that directly include the primary keywords from that ad group. This is crucial for relevance and Quality Score.
• -> Benefit-focused headlines: What's in it for the customer? "Save Time on Invoicing," "Get a Quote in 60 Seconds," "24/7 Emergency Service."
• -> Feature-focused headlines: What makes your product or service unique? "UK-Based Support Team," "Made with Organic Cotton," "Integrates with Xero."
• -> Trust-building headlines: Why should they believe you? "Over 10,000 Happy Customers," "Rated Excellent on Trustpilot," "Established in 1998."
• -> Call-to-Action (CTA) headlines: Tell them what to do. "Shop The Sale Now," "Book a Free Consultation," "Download Your Guide."

The same logic applies to your descriptions. Mix and match different angles. Your goal is to create a diverse set of messages so the algorithm can find winning combinations for different types of searchers.

4. Use Pinning Sparingly and Strategically: The 'pin' feature lets you force a headline or description to appear in a specific position (e.g., this headline MUST appear in position 1). It's tempting to use this to regain the control you had with ETAs, but be careful. Over-pinning defeats the purpose of an RSA. It restricts the algorithm's ability to test and learn. I generally reccomend:

• -> Pinning your brand name to position 1 or 2: This is often a good idea for brand consistency.
• -> Pinning a core, must-see message: If you have a critical piece of information like "Free UK Delivery," you might pin that to position 2 to ensure it's always visible.
• -> Leaving position 3 unpinned: This gives the algorithm a flexible slot to test your more creative or varied headlines.

A good starting point is to pin one headline to position 1, pin another to position 2, and leave everything else unpinned to give Google maximum flexibility. Below is a table that shows how you might structure the assets for a fictional cleaning service company.

You'll need to analyse the results correctly...

Once your test is live, the temptation is to check it every five minutes. Don't. You need to let it run and gather enough data to make a smart decision. Deciding the winner of a test based on a handful of clicks is a classic mistake.

1. Define Your Winner: Before you even look at the data, you need to know what you're looking for. What is the single most important metric for your business? For 99% of businesses, it’s not clicks or Click-Through Rate (CTR). These are vanity metrics. An ad can have a fantastic CTR but never convert, making it a waste of money.

Your winning metric should be one of these:

• -> Cost Per Conversion (or CPA - Cost Per Acquisition): This is usually the best one. How much are you paying for each lead or sale? The ad that delivers conversions for the lowest cost is often the winner.
• -> Conversion Rate (CVR): What percentage of people who click the ad go on to convert? This shows how effective the ad's message is at persuading the *right* kind of people to click.
• -> Return on Ad Spend (ROAS): If you're in eCommerce and can track revenue, this is the ultimate metric. For every £1 you spend on an ad, how many pounds do you get back?

Pick ONE of these as your primary success metric. You can look at the others as secondary indicators, but make your final decision based on your main goal.

2. Wait for Statistical Significance: How long should you run the test? The answer is "until it's statistically significant." This means you have enough data that the result is unlikely to be due to random chance. As a rule of thumb, I'd suggest letting each ad get at least a few hundred clicks and run for at least 1-2 weeks. If your budget is small, it might take longer. Don't call a winner until you have at least 10-20 conversions for each ad variant. The more data, the more confident you can be in your decision.

To help with this, I've built a simple calculator. You can plug in the spend and conversion numbers for your two ads (your old ETA vs your new RSA), and it will instantly calculate the CPA for each and show you the performance difference. This takes the guesswork out of comparing them and helps you spot the clear winner based on the metric that actually matters.

3. Iterate and Improve: Your analysis doesn't stop after you pick a winner. If the RSA wins, great! You can now pause the old ETAs in that ad group. But you can also make the RSA even better. Inside the Google Ads interface, you can view the asset performance report for your RSA. It will rate each headline and description as 'Low', 'Good', or 'Best'.

Your job is to systematically replace the 'Low' performing assets with new ideas. Treat it like a continuous improvement cycle. This is how you go from a good campaign to a great one. You don't just 'set and forget'; you constantly test and refine.

We'll need to look at the bigger picture: what really drives performance...

Finally, I want to take a step back. It's easy to get fixated on a specific tactic like switching ad formats. But in the grand scheme of things, the type of ad you use is a relatively small piece of the puzzle. If your campaigns aren't performing as well as you'd like, changing from ETAs to RSAs is like rearranging the deckchairs on the Titanic. It won't save a sinking ship.

True success in paid advertising comes from getting the fundamentals right. Based on my experience running campaigns for everyone from B2B SaaS companies to eCommerce stores, the things that *really* move the needle are almost always the same. I remember one campaign for a software client where we reduced their Cost Per User from £100 down to just £7, and it wasn't because we found a magic ad format. It was because we fixed the core strategy.

Here’s what really matters:

1. Your Offer: This is, without a doubt, the number one reason campaigns fail. Is your offer genuinely compelling to your target audience? An amazing ad can't sell a bad offer. You mention your campaigns are already performing well, which is great – it means your offer probably has some appeal. But could it be better? Are you clearly communicating a solution to an urgent, expensive problem? Are you making it easy and low-risk for people to take the first step? An offer like "Request a Demo" is high friction. An offer like "Start a Free Trial (No Credit Card)" is low friction. A new ad format will never be as impactful as improving your core offer.

2. Your Targeting: Are you reaching the right people? You can have the best ad and the best offer in the world, but if you show it to the wrong audience, it's completely useless. This means digging deep into your keyword strategy on Search. Are you targeting keywords with clear commercial intent (e.g., "emergency plumber london") rather than just informational queries (e.g., "how to fix leaky tap")?

3. Your Landing Page: The ad's only job is to get the right person to click. After the click, the landing page has to do all the heavy lifting. Does your landing page continue the 'scent' from the ad? Is the messaging consistent? Is it fast, mobile-friendly, and trustworthy? Does it have a single, clear call to action? I've seen countless campaigns with great ads fail because they point to a slow, confusing homepage instead of a dedicated, optimised landing page.

The ad format is just the vehicle. The offer, audience, and landing page are the engine, the fuel, and the destination. To put it in perspective, here's how I generally see the components of a successful campaign breaking down in terms of impact.

As you can see, the specific ad format falls into that last 10%. It's important to get it right, and testing is the way to do that. But don't expect it to revolutionise your results on its own. The real gains are found in optimising the bigger, more strategic pieces.

So, to bring it all together, my advice is to take a methodical, risk-averse approach. Your current success is valuable – protect it. Use clean A/B tests to challenge your assumptions and find incremental gains, rather than making sweeping changes based on what you *think* might happen.

I've detailed my main recommendations for you below:

This is the main advice I have for you, distilled into an actionable plan. Following this process will allow you to test flexible ads effectively without risking the strong performance you've already built.

This structured testing is exactly the kind of nuanced work that can be tricky to manage on your own, especially when you're busy running a business. It requires patience, a disciplined approach to data analysis, and a deep understanding of how the ad platforms really work underneath the hood. This is often where getting expert help can make a huge difference – turning good campaigns into market-leading ones.

If you'd like to chat through your specific campaigns and have a second pair of expert eyes look over your account, we offer a completely free, no-obligation initial consultation. We could review your current setup together and identify the biggest opportunities for growth, whether that's through ad format testing or those bigger-picture strategic elements.

Hope this has been helpful and gives you a clear path forward!

Regards,

Team @ Lukas Holschuh