Most B2B Google Ads accounts we audit are not losing money because of bad keywords or weak ad copy. They are losing it because the bid strategy is misconfigured for the sales cycle, the conversion data feeding Smart Bidding is polluted, or the account was moved to an automated strategy before it had enough signal to work with. Each of these mistakes can independently push your cost-per-lead 40-80% above what it should be, and they compound when they occur together.
Switching to Smart Bidding Too Early
Target CPA and Maximize Conversions both require a statistically meaningful conversion volume before Google's auction-time bidding model can make reliable predictions. Google's own guidance recommends at least 30-50 conversions per month at the campaign level before switching, and in practice, B2B campaigns with long sales cycles rarely hit that threshold in the first 60-90 days. When you force Smart Bidding on a thin-data campaign, the algorithm fills the gap with guesses, and those guesses tend to favour cheaper, higher-volume clicks rather than the specific intent signals that produce qualified pipeline.
The correct sequence is to start on Maximize Clicks with a manual CPC cap, build 6-8 weeks of conversion history, and then migrate to Target CPA with a starting target set 20-30% above your actual average CPL so the algorithm has room to learn without immediately restricting volume. Cutting that target too aggressively on day one of Smart Bidding is one of the fastest ways to stall impression share and push CPLs upward as the system chases an unachievable number.
Feeding Smart Bidding the Wrong Conversion Actions
Smart Bidding optimises toward whatever conversion action you point it at. If your primary conversion is a contact form submission that includes both enterprise prospects and low-intent competitors checking your pricing, the algorithm will optimise for volume of that mixed signal, not for quality. We regularly see accounts where 30-40% of form fills are from non-ICP sources, which means the bidding model is partially trained on the wrong behaviour. The fix is to use Google Ads conversion value rules or separate campaigns by funnel stage so that high-intent actions like demo requests or pricing page visits carry more weight than generic contact submissions.
One approach that works well in B2B is importing offline conversion data from your CRM, specifically qualified opportunity or SQL status, back into Google Ads with a 30-60 day delay. This trains the algorithm on real pipeline signals rather than raw form volume. It takes more setup work, but accounts that implement this consistently see CPL drops of 25-45% within a single quarter because the bidding model starts prioritising the traffic segments that actually convert downstream. For more on how attribution affects this decision, see our breakdown of multi-touch attribution for B2B ROI.
Ignoring Auction Insights and Position Data
Many B2B accounts run Target Impression Share strategies without defining where in the auction they actually need to appear. Targeting 90% impression share at the absolute top of page for a high-volume keyword like "project management software" will burn budget competing against Asana and Monday.com's seven-figure monthly ad spend. The smarter play for most mid-market B2B advertisers is to target 60-75% impression share across the full page, which preserves auction competitiveness without forcing every bid to its ceiling just to win position one.
Check your Auction Insights report monthly. If competitors with significantly higher domain authority and budget are consistently outranking you at position one, the answer is not to raise bids, it is to tighten your keyword list to the segments where your offer has a genuine conversion advantage and let the broader terms go. This is directly related to how you have structured your campaign hierarchy in the first place, and if that foundation is weak, no bid strategy will save it. Our guide on how to structure Google Ads for B2B covers the segmentation logic in detail.
Common Bid Strategy Mistakes at a Glance
- Activating Target CPA before accumulating 30+ monthly conversions at campaign level
- Setting the Target CPA below the actual historical average CPL from day one
- Including low-quality or mixed-intent form fills as primary conversion actions
- Using Target Impression Share without defining a position cap or budget ceiling
- Applying a single bid strategy across campaigns with fundamentally different funnel stages
- Never importing CRM offline conversions to refine Smart Bidding signal quality
Portfolio Bid Strategies and When They Backfire
Portfolio bid strategies, where you group multiple campaigns under a shared Target CPA or Target ROAS, can smooth out conversion volatility across a large account. The problem is that B2B campaigns with very different average deal sizes, sales cycles, or audience segments should almost never share a portfolio strategy. A brand campaign converting at a $40 CPL and a competitor campaign converting at $180 CPL will average out to something that satisfies neither objective. The algorithm will redistribute budget toward the easier conversions, which are typically brand terms, and starve the prospecting campaigns that are doing the actual growth work.
Separate your portfolio strategies by funnel intent: one for brand and retargeting campaigns, one for high-intent non-brand campaigns, and a third if you run display or Performance Max alongside search. This gives you clean performance data per segment and prevents the algorithm from masking inefficiency through averaging. If your CPLs are consistently higher than industry benchmarks, the root cause is often structural like this rather than a keyword or copy problem. For a direct comparison of where your numbers should sit, our Google Ads B2B costs and benchmarks article has current figures by industry and region.
The Learning Period Trap
Every time you make a significant change to a Smart Bidding campaign, including adjusting the Target CPA by more than 15-20%, adding a new conversion action, or restructuring ad groups, you reset the learning period. During learning, Google's own documentation notes that performance can be unpredictable for 1-2 weeks. B2B accounts that make frequent changes in response to short-term CPL spikes end up in a permanent learning loop where the algorithm never stabilises. The CPL looks volatile and keeps climbing, but the actual problem is the constant intervention, not the strategy itself.
The discipline required here is to make one structural change at a time, wait a full two-week learning window, and evaluate performance only after that window closes. Set a decision rule before you touch anything: define the CPL threshold and the observation period in advance so you are not reacting to a single bad week. This operational discipline is unglamorous, but it is what separates accounts that hit $120 CPLs from accounts that are stuck at $340 for the same keyword set and the same budget.