A founder running a B2B software company in Bangalore switched her Google Ads account to Performance Max on the advice of her agency. Within three months, impressions were up 340%, spend was up 60%, and qualified leads were down 22%. The agency sent a report showing strong 'conversion volume.' She dug into the data and found that most of those conversions were people spending more than two minutes on the website. Not leads. Not enquiries. Time-on-site.
This is not an unusual story. Performance Max, Google's AI-driven campaign type that allocates budget across Search, Shopping, Display, YouTube, Discover, Gmail, and Maps simultaneously, is genuinely powerful when configured correctly and actively managed. It is also one of the easiest campaigns to mismanage, because the defaults favour Google's objectives (maximum conversion volume) rather than yours (qualified leads at an acceptable cost).
This guide is written for business owners and marketing decision-makers who want to understand what Performance Max is doing with their budget, why the default configuration often underperforms for lead generation, and what specific controls are available to take back meaningful oversight. The guide does not recommend avoiding PMax. It recommends running it with your eyes open.
Performance Max gives Google significant latitude to decide where your ads appear and who sees them. The question is not whether to use it. It is how much of that latitude is justified by the results it is producing.
Performance Max is Google's fully automated campaign type, launched broadly in 2022 and now the recommended format for most Google Ads objectives. It runs across all of Google's owned inventory from a single campaign: Google Search (including Search Partner networks), Google Shopping, Display Network, YouTube, Discover feed, Gmail, and Google Maps. The algorithm decides which channels, audiences, placements, and creative combinations to use based on the conversion goal you define.
The core promise of PMax is efficiency through automation: Google's machine learning can optimise across more signals and more inventory simultaneously than a human campaign manager can, in theory producing better results for the same budget. In practice, that promise holds reasonably well for e-commerce businesses with large product catalogues, clear purchase conversion data, and high transaction volumes that give the algorithm the data it needs to learn quickly.
For lead generation businesses, professional services, B2B companies, and low-volume conversion categories, the promise holds less consistently. The algorithm optimises for conversion volume, not conversion quality. If form submissions are the tracked conversion and the business receives a mix of highly qualified and completely unqualified form submissions, PMax will optimise for quantity without distinguishing between the two. It cannot know that a lead from a specific job title in a specific company size is worth ten times more than a generic enquiry, unless that signal is fed back to it deliberately.
Most PMax campaigns are launched with default settings and a broad conversion goal. That combination is the source of most of the performance problems business owners encounter.
Google's algorithm optimises for whatever conversion action is defined in the account. The most common mistake in lead generation PMax campaigns is using a weak conversion signal: page visits over a time threshold, scroll depth, or generic form starts. The algorithm will generate plenty of these. They are not leads. A qualified lead requires a specific, high-intent action: a completed enquiry form, a phone call of meaningful duration, a booked appointment, or a product demo request. Every PMax campaign for lead generation should be optimising for the conversion that most closely maps to a qualified sales-ready lead.
PMax allows advertisers to provide audience signals: lists and characteristics that tell Google's algorithm who the best customers look like. Without these signals, the algorithm starts from scratch and spends significant budget in the learning phase reaching a broad, unfocused audience. With strong audience signals (a CRM upload of existing customers, website visitor lists, customer match data), the algorithm has a starting point that compresses the learning phase and concentrates early spend on higher-probability audiences.
Most PMax campaigns launched without agency involvement skip this step entirely. Campaigns launched with agency involvement often use generic interest-based signals rather than first-party data. The difference in lead quality between a PMax campaign with a strong CRM-based audience signal and one with no signal is significant enough to justify treating it as a prerequisite, not an optional enhancement.
PMax campaigns automatically bid on branded keywords, meaning searches for the company's own name. Branded searches typically have very high conversion rates because the user already knows the business and is actively looking for it. PMax claims these conversions, inflating its reported performance, while the budget that should be serving non-branded prospecting searches gets partially diverted to converting users who would have found the business anyway through organic search.
The solution is to add brand terms as negative keywords at the account level and run a separate, lower-budget branded Search campaign to handle those queries efficiently. This separation keeps PMax focused on new audience acquisition rather than harvesting existing brand demand.
PMax's Display and YouTube inventory includes a large amount of low-quality placement: mobile game apps, thin content websites, and YouTube pre-rolls that users skip in the first five seconds. The algorithm allocates budget across all available inventory without quality filtering unless placement exclusions are applied. For B2B and professional service advertisers, a significant portion of PMax Display spend often lands on placements that will never produce a qualified lead regardless of how well the ad is written.
The table below maps the key Performance Max optimization levers available to advertisers, the impact each one has on lead quality and budget efficiency, and the complexity of implementing each control.
| Optimization Control | What It Does | Impact on Lead Quality and Spend | Implementation Complexity |
| Conversion action definition | Tells Google's algorithm what to optimise for. Strong signal means specific high-intent actions like completed forms or phone calls. | High. The single most important configuration decision. Weak conversion signals produce high volumes of low-quality interactions that inflate reported performance. | Low. Set in Google Ads conversion tracking before campaign launch. Requires correct tracking implementation. |
| Audience signals (CRM upload and customer match) | Provides first-party data to seed the algorithm's audience learning. Customers, qualified leads, and website visitors. | High. Compresses learning phase. Directs early budget toward proven audience profiles rather than broad exploration. | Medium. Requires CRM export, formatting to Google's requirements, and customer match eligibility (minimum list size and account history). |
| Brand keyword exclusions | Prevents PMax from bidding on branded search terms, redirecting budget to new audience acquisition. | Medium to High. Stops PMax from claiming easy branded conversions that inflate performance metrics while reducing genuine prospecting efficiency. | Low. Account-level negative keyword list. Requires identifying all brand term variants including common misspellings. |
| Placement exclusions (mobile apps and low-quality sites) | Removes specific placement categories from Display and YouTube inventory. | Medium. Reduces wasted Display spend on mobile gaming apps and thin content sites that generate impressions without converting. | Medium. Requires building and maintaining exclusion lists. Google provides category-level exclusions for mobile apps. |
| Asset group segmentation by audience and intent | Creates separate asset groups with different creative and messaging tailored to different audience segments or funnel stages. | Medium to High. Allows the algorithm to test creative relevance against specific audience segments rather than running generic creative across all placements. | Medium to High. Requires creative production for each segment. B2B campaigns benefit from separating by company size, industry, or job function. |
| Target ROAS or Target CPA bidding with realistic floors | Sets cost efficiency constraints that prevent Google from buying cheap, low-quality conversions at scale. | High for lead quality. A Target CPA slightly above the current average CPA focuses spend on the placements and audiences that convert most efficiently rather than maximising volume. | Medium. Requires sufficient conversion data before switching (minimum 30 to 50 conversions in the past 30 days for reliable algorithm performance). |
| Search term insight monitoring | Reviews the search category insights available in PMax to understand which query themes are driving conversions. | Medium. Google does not provide individual search terms for PMax, but category-level insights reveal whether traffic quality matches intent. Informs negative keyword decisions. | Low. Available in the Insights tab of the PMax campaign. Requires weekly review habit rather than technical setup. |
Table: Performance Max optimization controls ranked by impact, with implementation complexity and lead quality effect
The most effective Google PMax strategy for lead generation is not about fighting the algorithm. It is about giving the algorithm the right inputs so its optimisation works in the direction of business outcomes rather than against them. The following sequence represents the practical setup order for a PMax campaign being rebuilt or launched with lead quality as the primary objective.
PMax reporting is deliberately limited compared to standard Search campaigns. Google does not provide placement-level or search term-level detail for PMax. What it does provide is usable if reviewed consistently with the right questions.
The most important weekly check for lead generation PMax campaigns is not how many conversions the campaign reported. It is how many of those conversions became qualified sales conversations. Track this by connecting Google Ads conversion data to CRM records: how many PMax-attributed form submissions resulted in a sales rep calling a real prospect? What percentage of PMax leads were marked as qualified in the CRM? If conversion volume is rising but qualified lead rate is falling, the algorithm is finding cheaper conversions, not better ones.
Check asset group performance weekly. If one asset group is receiving 80% of impressions and another is receiving almost none, investigate why. Either the high-impression group is genuinely outperforming or the algorithm has settled into a pattern based on early data that may not reflect long-term performance. Asset groups that receive very few impressions cannot be fairly evaluated and may need creative refreshing or audience signal adjustment.
PMax provides a channel breakdown showing what proportion of budget went to Search, Shopping, Display, YouTube, and other channels. For B2B lead generation, a campaign spending 60% of its budget on Display and YouTube with poor conversion rates from those channels has a rebalancing problem the advertiser can address by using asset group configuration to de-emphasise visual channels. A campaign spending most of its budget on Search-type queries is typically performing more predictably for lead quality.
PMax is not universally appropriate. There are specific scenarios where a standard Search campaign, a targeted Display campaign, or a combination of the two will outperform PMax for the same budget.
The most advanced but most impactful technique for improving PMax lead quality is closing the signal loop between the CRM and Google Ads. This is the practice that separates businesses achieving profitable ad spend from those generating impressive impression counts and disappointing pipeline.
The standard setup tracks form submission as conversion and stops there. Google sees that a user clicked an ad and completed a form. It does not know whether that user became a qualified lead, a paying customer, or an irrelevant enquiry. Every completed form looks the same to the algorithm.
The closed-loop setup uses Google Ads' offline conversion import feature to feed CRM outcomes back into the algorithm: when a lead is marked as qualified in the CRM, that outcome is uploaded back to Google Ads and associated with the original click. When a lead converts to a paying customer, that outcome is uploaded with its revenue value. The algorithm now has signal quality data: it learns that certain audience characteristics, certain search themes, and certain placement types produce leads that become customers, rather than just leads that complete forms.
This setup requires CRM integration work, consistent data upload processes, and sufficient volume to be statistically meaningful. For businesses generating 15 or more qualified leads per month, the investment in closing the signal loop produces measurable improvement in PMax lead quality within two to three months of implementation.
Bud is a creative and digital marketing agency based in Bangalore, operating since 2010 across real estate, healthcare, FMCG, B2B, education, and lifestyle categories. As a Google Premier Partner, Bud manages Google Ads accounts across Search, Shopping, Display, YouTube, and Performance Max for brands across South India. The Google Premier Partner status means Bud has access to Google's beta features, dedicated support, and benchmarking data across managed accounts.
The PMax management approach at Bud follows the principle that automation works best when the inputs are correct and the measurement is honest. Every PMax engagement starts with a conversion audit: what is currently being tracked, whether it maps to a genuine business outcome, and whether the reported performance would hold up if evaluated against CRM data rather than platform metrics. A campaign showing a Rs. 400 cost per conversion that produces Rs. 8,000 cost per qualified lead is not performing at Rs. 400. It is performing at Rs. 8,000. The audit makes that visible before the strategy is built.
When a business approaches Bud as its Google ads agency, the PMax strategy is built from the commercial objective backward: what does a qualified lead look like, what is the acceptable cost per qualified lead, and what measurement infrastructure exists or needs to be built to track that outcome through the funnel. The creative brief, audience signal build, asset group structure, and bidding strategy all follow from those commercial answers.
Bud has won two Gold and three Silver at the Big Bang Awards 2025. When a brand needs a PPC Agency that manages Google Ads as a commercial growth programme rather than a media buying function, the starting point is always the same: what does the business actually need to grow, and how does the ad spend connect to that outcome. Bud has worked on paid media programmes at scale for real estate developers, healthcare groups, FMCG brands, B2B tech companies, and D2C brands across South India, and the Performance Max work sits within that broader programme rather than as an isolated campaign type.
No. PMax works best alongside a carefully maintained branded Search campaign and, where applicable, a Shopping campaign for e-commerce. Replacing all campaign types with PMax removes the segmentation and control that branded campaigns provide and can result in brand keyword cannibalisation with no separation to manage it. The recommended structure is PMax for non-branded prospecting, a dedicated branded Search campaign for brand term queries, and Shopping campaigns retained separately if the Google Merchant Center feed is used for e-commerce.
Almost always, the conversion definition is the problem. If the campaign is optimising for a weak conversion signal (time on site, page visits, scroll depth, or form starts), it will produce many of those weak signals. The algorithm is doing exactly what it was told to do. The fix is to change the primary conversion action to a genuine high-intent signal and give the algorithm time (four to six weeks) to re-optimise against the new goal.
Run a geo-based holdout test: pause PMax in a comparable market or geography for four to six weeks while running it normally in the primary market. Compare conversion rates and organic search volume between the two. If PMax is generating incremental conversions beyond what organic would have produced, the incremental volume will be visible in the comparison. If organic performance is identical in both geos, PMax may be claiming credit for demand it did not generate. This test is the most reliable way to evaluate incrementality without full-account measurement infrastructure.
As a rough guide, PMax needs sufficient budget to generate at least 30 to 50 conversions per month for the algorithm to optimise meaningfully. For most Indian B2B lead generation categories, this typically means a minimum monthly budget of Rs. 80,000 to Rs. 1,50,000 depending on the CPC levels in the category. Below that threshold, a targeted Search campaign with manual or enhanced CPC bidding gives more control over the limited data available and produces more predictable results while conversion history accumulates.
Performance Max optimization is not about disabling automation. It is about giving the automation the right signals, the right constraints, and the right measurement framework to optimise toward business outcomes rather than platform metrics. A PMax campaign with a strong conversion definition, first-party audience signals, brand exclusions, segmented asset groups, and CRM signal loop will outperform both a poorly configured PMax campaign and most manually managed campaign structures for the same budget.
The businesses that achieve profitable ad spend from PMax are not the ones that trusted the defaults and hoped for the best. They are the ones that treated the campaign as a strategic input-output system: quality inputs (conversion goals, audience signals, creative) produce quality outputs (qualified leads at an acceptable cost). Poor inputs produce impressive dashboards and thin pipelines.
Taking back control of PMax does not mean removing the automation. It means configuring the automation to serve the business rather than the platform. That is a different objective from what the default settings assume, and achieving it requires deliberate decisions at every stage of the campaign setup and ongoing management process.
Google Performance Max will optimise for whatever you tell it to. The problem is that most businesses tell it to optimise for the wrong thing and then wonder why the results do not match the business goals.
Bud India | Creative Advertising Agency, Bangalore
WE ARE AN OFFICIAL GOOGLE PREMIER PARTNER
