d2b2 brings data science thinking to paid marketing and revenue operations. We audit your stack, build the systems, teach your team how they work, and hand you the keys when the engagement ends. Services include paid marketing setup, revenue analytics, marketing infrastructure, and answer engine optimization.

What is the d2b2 engagement process?

Every d2b2 engagement follows three phases. The Audit: a structured review of your stack, spend, and data to understand what exists and what needs fixing. The Stack Build: implementing tracking, attribution, AI tools, and automation that work correctly from day one. The Handoff: full documentation so your team can run everything independently after the engagement ends.

Who does d2b2 work with?

d2b2 works with small businesses and growth-stage companies where marketing has historically been reactive and underpowered. All engagements are project-based, not retainer-based. Based in San Diego, CA, serving clients across the US.

How is d2b2 different from a traditional marketing agency?

d2b2 applies data science thinking to marketing rather than relying on intuition or waiting on monthly reports. The goal is to build systems you own and understand, not dependency on an ongoing agency relationship. Every engagement ends with a complete handoff so the client can run the work independently.

Why Your PMAX Campaigns Aren't Converting (And What to Fix)

Performance Max campaigns promise a lot: one campaign, all of Google’s inventory, machine learning doing the optimization. In practice, PMAX underperforms for a lot of advertisers, not because the product is broken, but because it requires better inputs than most accounts provide.

If your PMAX campaigns aren’t converting at the rates you expected, here’s where to look.

Ad Rank is still the lever

A common misconception is that PMAX replaces the auction. It doesn’t. PMAX still competes in auctions across Search, Display, YouTube, Gmail, and Discover, and Ad Rank still determines when and where your ads show.

Ad Rank is a function of bid, Quality Score, expected CTR, ad relevance, landing page experience, and auction-time signals. PMAX doesn’t remove these factors; it automates bidding decisions across them. If your Ad Rank is weak, PMAX will show your ads in lower-value placements and lose competitive auctions.

The practical implication: asset quality matters more in PMAX than in standard campaigns because your assets are what Google uses to assemble ads across formats. Poor assets mean poor expected CTR, which means poor Ad Rank.

Check your asset groups

Asset groups are where most PMAX problems originate. Each asset group should be:

Thematically tight. Don’t put “residential plumbing” and “commercial HVAC” in the same asset group. The algorithm uses assets to infer who to target. Mixed themes produce mixed signals.

Fully populated with high-quality assets. Google will tell you your asset strength: Poor, Good, or Excellent. Aim for Excellent. This means 15 headlines (all different, all specific to the theme), 5 descriptions at full character count, multiple high-quality images in the required aspect ratios, and at least one video. If you don’t provide a video, Google will generate one from your images, and it will not represent your brand well.

Aligned to a specific audience signal. Audience signals don’t restrict who sees your ads, but they give the algorithm a starting point. Use your customer list, website visitors, or in-market segments that match the product or service in the asset group.

Conversion tracking must be accurate

PMAX is aggressive about optimizing toward its conversion targets. If your conversion tracking counts low-quality signals, such as contact page visits, time on site, or soft engagements, as primary conversions, PMAX will optimize toward those instead of actual business outcomes.

Audit what’s set as a primary conversion action before running PMAX. In most cases, primary conversions should be things with clear revenue implications: form submissions from qualified leads, phone calls over a certain duration, actual purchases.

If you have both lead form conversions and purchase conversions, consider separate campaigns with different optimization goals.

Give it time, but with a floor

PMAX needs a learning period, typically 6 to 8 weeks with sufficient conversion volume. During this period, performance will fluctuate as the algorithm builds frequency and conversion data.

However, “give it time” has limits. If a campaign has been running for 8 weeks with low impressions and near-zero conversions, something structural is wrong. Usually it’s budget constraints, targeting signals that are too narrow, or conversion tracking that isn’t firing.

A reasonable threshold: if PMAX hasn’t produced 20 or more conversions in the first 6 weeks, pause it and review the inputs before restarting.

Check the Insights tab

The Insights tab in PMAX shows which audience segments, search themes, and creative assets are driving results. This is one of the few places PMAX gives you visibility into what’s actually happening.

Look for:

Search themes that are converting (add them to your standard Search campaigns if you want more control)
Asset performance ratings (replace Poor-rated assets immediately)
Audience segments that are performing unexpectedly well or poorly

The bigger picture

PMAX performs best when you give it good data, good assets, and a realistic conversion target. The algorithm is powerful, but it’s not a replacement for strategy. Accounts that see PMAX fail usually haven’t invested in the inputs.

d2b2 specializes in paid campaign architecture including PMAX setup and optimization. If your campaigns aren’t delivering, start a conversation.