I've been doing this for a while. Boutique shops, holding companies, brands you'd recognize on the shelf — and now a decade running my own agency. Across all of it, I've personally managed or overseen well north of $50 million in ad spend.
And I keep seeing the same mistakes. Over and over. In accounts spending $5K/month and accounts spending $500K/month. In accounts managed by in-house teams and accounts managed by agencies that should know better. These seven things show up so consistently that at this point I could almost run an audit with my eyes closed.
I'm not writing this to be mean about whoever managed your account before. Most of these mistakes are understandable — they come from being busy, from trusting the platform too much, or from simply not knowing what you don't know. But they're costing you real money. So let's get into it.
1. Not Using Negative Keywords (Yes, Still, in 2026)
I audited an ecommerce account late last year — mid-six-figures in annual Google Ads spend — and they had zero negative keywords. None. Not a single one across any campaign.
I pulled the search term report for the previous 90 days and started highlighting irrelevant queries. "Free [product]." "[Product] DIY tutorial." "[Competitor name] careers." Random informational queries that had nothing to do with buying intent. When I added it all up, roughly 23% of their spend was going to search terms that had virtually zero chance of converting. On their budget, that was about $4,200 a month just... gone. Lighting money on fire would've been more entertaining.
The fix is almost embarrassingly simple: build a negative keyword list. Review your search terms weekly (or at minimum biweekly). Add irrelevant terms. Build shared negative keyword lists at the account level for the obvious stuff — "free," "jobs," "salary," "DIY," "reddit" — and then build campaign-specific lists based on what you see in the data.
This is PPC 101 and I still find it missing in probably 60% of the accounts I audit. It's not glamorous work. Nobody's writing LinkedIn posts about their negative keyword strategy. But it's the equivalent of plugging the holes in a bucket before you pour more water in. Do it first.
2. Trusting Google's Automated Recommendations Blindly
Google really, really wants you to click that "Apply" button on their recommendations. They even gave you an optimization score to make you feel bad if you don't.
Here's the thing: Google's recommendations are optimized for Google's revenue, not your profitability. Some of them are fine — "add responsive search ad assets" is usually reasonable. But others are actively harmful. "Raise your budget" on a campaign that's already unprofitable? "Add broad match keywords" when your negative keyword list is nonexistent? "Enable auto-apply for ad suggestions" so Google can write your ad copy for you?
I took over an account once where the previous manager had turned on auto-apply recommendations. ALL of them. Google had been automatically adding broad match keywords, raising budgets, and even creating new ad copy without any human review. The account was a mess. Keywords that had nothing to do with the business. Budget increases on underperforming campaigns. Auto-generated ads that were grammatically correct but strategically nonsensical.
The optimization score was 97%. The actual ROAS was tanking.
The fix: Treat the optimization score as what it is — a vanity metric designed to get you to spend more. Review recommendations manually. Apply the ones that align with your actual strategy. Dismiss the ones that don't. And for the love of everything, turn off auto-apply. Go check right now: Settings > Auto-apply. If anything is toggled on that you didn't explicitly choose, turn it off today.
3. Ignoring Search Term Reports
This is related to negative keywords but it's actually a bigger issue. The search term report is the single most valuable report in Google Ads, and most people check it... never? Maybe quarterly if they're feeling ambitious?
Your search term report tells you exactly what people typed before clicking your ad. Not what keywords you're bidding on — what people are actually searching for. With broad match and phrase match, there can be a massive gap between those two things. I've seen brand campaigns triggered by competitor names. I've seen "luxury [product]" campaigns triggered by "cheap [product] alternative." The algorithm does its best, but it's not psychic, and it gets things wrong constantly.
Beyond the negative keyword opportunities, search term reports are a goldmine for positive discoveries. You'll find long-tail queries you never thought of that convert beautifully. You'll find patterns in how your customers describe their problems that can inform your ad copy and landing pages. One of our team members found a search term pattern for a B2B client that revealed an entire market segment they hadn't been targeting — a use case for their product that the client hadn't even considered. That discovery turned into their highest-performing campaign within two months.
The fix: Check search terms weekly. I mean it. Block 30 minutes on your calendar every Monday. Add negatives, identify new keyword opportunities, and look for patterns. If you're spending more than $10K/month on search and not reviewing search terms weekly, you are actively choosing to not know where your money goes.
4. No Conversion Value Tracking (or Broken Tracking)
OK so this one makes me genuinely a little heated because the consequences are so severe and so invisible.
Smart bidding — Target ROAS, Maximize Conversion Value, all of it — relies on your conversion data to make bidding decisions. If your conversion tracking is wrong, smart bidding is optimizing toward garbage. And it will do so very efficiently, which makes it even more dangerous. Your CPA might look great while your actual business results are terrible.
I audited an account last year that was reporting a 4.2x ROAS in Google Ads. The client was confused because their actual revenue numbers didn't match. Turns out, their conversion tracking was double-counting purchases — a Google tag and a Google Tag Manager tag were both firing on the same thank-you page. Every sale was being recorded twice. Actual ROAS was 2.1x. The smart bidding algorithm had been making decisions based on inflated data for months, which meant it was bidding aggressively on lower-quality traffic that appeared profitable but wasn't.
Another common version: no conversion value at all. Every conversion counted as equal. A $12 impulse buy weighted the same as a $400 high-margin product. The algorithm was drowning in cheap conversions on low-AOV products while the high-margin stuff got starved of budget.
The fix: Audit your conversion tracking. Actually open Google Tag Assistant, walk through your purchase flow, and verify that each conversion fires exactly once with the correct value. If you're running B2B, make sure you've set up conversion values that reflect actual lead quality — not all form fills are equal, and your bidding should know that. If you're running ecommerce, verify that dynamic revenue values are passing correctly. Do this quarterly at minimum, and always after any website changes.
5. Running the Same Creative for 6+ Months
This one shows up more on Meta than Google, but it absolutely applies to both.
On Meta, ad fatigue is a real and measurable thing. Your audience sees the same image or video enough times, they stop noticing it. CTR declines. CPA creeps up. The algorithm compensates by finding cheaper (usually lower-quality) placements. Performance slowly degrades, and because it happens gradually, nobody sounds the alarm until it's fallen off a cliff.
I took over a Meta account for a DTC skincare brand where the same three ad creatives had been running for eight months. EIGHT MONTHS. The frequency on their core retargeting audience was over 14 — meaning the average person in that audience had seen these ads fourteen times. Click-through rate was 0.3%. The account manager's explanation? "These were our best performers." Yeah. Six months ago. Now they're wallpaper.
On Google, people think this doesn't apply because it's search — the user has intent, so creative matters less. Not true. RSA (Responsive Search Ad) asset testing matters enormously. Google rotates your headlines and descriptions, but if you haven't added new assets in months, you're missing opportunities to test messaging angles that could improve CTR and conversion rate. I've seen accounts with RSA assets that all say basically the same thing — five headlines that are minor variations of each other. That's not testing. That's redecorating the same room.
The fix: On Meta, refresh creative every 3-4 weeks for prospecting audiences, 4-6 weeks for retargeting. Not necessarily entirely new concepts — sometimes a new format (static vs. video vs. carousel), new hook in the first 3 seconds, or new angle on the same offer is enough. On Google, review RSA asset performance monthly and swap out underperforming headlines and descriptions. Actually test meaningfully different messaging, not just synonyms.
6. Over-Segmenting Campaigns in the Smart Bidding Era
This one is the ghost of 2018 best practices haunting 2026 accounts.
There was a time — and I lived through it — when the right move in Google Ads was to segment everything. Separate campaigns by match type. By device. By geographic region. By time of day. The more granular your campaign structure, the more control you had over bids, and that control mattered because manual bidding rewarded precision.
Smart bidding flipped the script. Machine learning algorithms need data volume to work. They need enough conversions per campaign to learn patterns and make accurate predictions. Google's own guidance says 30+ conversions per month per campaign for Target CPA, and 50+ for Target ROAS. But I regularly see accounts with 15+ campaigns, each getting maybe 3-8 conversions a month. The algorithm is starved. It can't learn. It makes bad bidding decisions because it doesn't have enough signal.
We inherited a lead gen account that had 22 campaigns. Twenty-two. Total monthly conversions across the account: about 180. Sounds OK in aggregate, but 14 of those campaigns were getting fewer than 10 conversions each. We consolidated down to 6 campaigns. Same keywords, same ads, same budgets — just restructured so each campaign had meaningful conversion volume. Within 60 days, cost per lead dropped 28% and lead quality (measured by SQL rate) actually improved.
The fix: Count your conversions per campaign per month. If any campaign is consistently below 30, it's a candidate for consolidation. Yes, you lose some granular control. That's the trade-off, and in 2026, it's almost always worth it. Smart bidding with enough data beats manual optimization on starved campaigns every time. Use ad groups, audience signals, and labels to maintain the organizational clarity you need without fragmenting the conversion signal.
7. No Measurement Plan
This is the big one. The one that enables all the others.
"We track conversions" is not a measurement plan. It's a starting point. A measurement plan answers: What counts as a conversion? How do we value different conversion types? What attribution model are we using, and why? What's our conversion window? How do we handle cross-device and cross-channel attribution? What are our micro-conversions (leading indicators) versus macro-conversions (actual business outcomes)? How do we reconcile what the ad platform reports with what actually shows up in our CRM or revenue system?
For B2B, the measurement gap is especially painful. Google Ads reports a form fill as a conversion. Great. But was that form fill from a qualified buyer who turned into a $50K deal, or a student doing research? Without offline conversion import — where you feed back actual sales outcomes from your CRM to Google Ads — your smart bidding algorithm treats both identically. It optimizes for volume of form fills, not quality. I've seen B2B accounts where 70% of "conversions" were junk leads, and the algorithm was happily finding more of them because nobody told it the difference.
For ecommerce, the gap shows up differently. Maybe you're tracking purchases but not accounting for returns. Your reported ROAS is 5x but after returns it's 3.2x, and the campaigns driving the highest ROAS also have the highest return rate because they're attracting impulse buyers. Without feeding return data back into your measurement, you can't see this.
A real measurement plan also includes regular reconciliation. Once a month, sit down and compare what Google/Meta reports versus what your actual bank account or CRM says. There is always a gap. Understanding the size and nature of that gap is essential for making smart decisions about where to invest more and where to pull back.
The fix: Build a measurement plan before you spend a dollar on ads. Document your conversion actions, their values, your attribution model, your conversion window, and how you'll validate platform data against real business outcomes. For B2B: set up offline conversion import. I know it's a pain. Do it anyway — it's the single highest-impact thing you can do for lead quality. For ecommerce: make sure you're tracking revenue accurately and consider feeding back return data or LTV data if you can. Review your measurement plan quarterly because platforms change attribution defaults, conversion windows expire, and tags break.
The Throughline
If you read through all seven of these, you might notice a pattern. None of them are about creativity. None of them are about having some brilliant strategic insight or discovering a secret hack. They're all about doing the boring, foundational work that separates accounts that coast from accounts that compound.
Negative keywords aren't exciting. Auditing your conversion tracking is tedious. Building a measurement plan is a spreadsheet exercise, not a brainstorm. But after $50M+ in managed spend, I can tell you with absolute confidence: the accounts that perform best are not the ones with the cleverest strategies. They're the ones where someone is consistently doing the unglamorous work. Checking the search terms. Refreshing the creative. Verifying the data.
Every week. Every month. Without fail.
The spicy, attention-grabbing stuff — the new campaign launch, the creative concept that goes viral, the clever audience hack — that's the seasoning. The boring stuff is the whole meal. And most accounts I see are all seasoning, no substance. A beautifully structured campaign built on broken tracking and ignored search terms is just a very organized way to waste money.
So here's my challenge to you. Pick one of these seven. Just one. Go check your account right now. If you find the problem, fix it this week. Then next week, check the next one. By this time next month, you'll have caught most of the money you're currently leaving on the table.
And if you go through all seven and everything looks clean? Call me. I want to meet whoever's running your account, because that person is rare and you should never let them go.