The ghost effect: why 40% of your conversions don't show up for 90 days

Here's something that happens in every B2B account I've ever audited, and almost nobody talks about it.

You're looking at your Google Ads conversion report for March. You see 47 conversions. You celebrate. You report it up. You tell your team to keep doing what they're doing in March. Then May rolls around and suddenly your March numbers get revised upward to 78 conversions. Then July, they're at 94.

That's the ghost effect. Roughly 40% of your conversions are arriving late, hidden, buried in the fog of a long sales cycle. And if you're making optimization decisions based on what you see on day 30, you're making them blind.

Why this happens

The short version: long sales cycles break attribution.

In a SaaS signup flow, the entire user journey takes maybe 15 minutes. Email to landing page to form to confirmation. Your pixel fires, Google records a conversion, everyone knows what happened. Latency is near zero. The data is clean.

In enterprise B2B, the journey looks different. A procurement engineer finds your site in January. They spend 45 minutes reading specs, checking pricing, checking your case studies. Then they leave. Radio silence for four weeks. February, they come back, spend an hour, leave again. In March, they call your sales team directly. No click. No pixel. No conversion record.

By April, the deal is in Salesforce as a real opportunity. By July, it closes.

In your Google Ads account, this shows up as... nothing. Or it shows up 89 days later when your CRM finally connects back to Google and says "Hey, remember that click on January 15th? It closed today."

Or if you're not that sophisticated, it never shows up at all.

The ghost effect is the gap between when the conversion actually occurred in your sales process and when Google records it. In most B2B accounts, that gap is enormous.

What this does to your optimization

Here's where it gets dangerous.

You're running a campaign in January. It generates 20 clicks. Your conversion rate is terrible. 0%. You look at the data, decide the campaign sucks, and pause it.

What you don't know is that three of those clicks converted. They just haven't shown up yet. You won't see them until April or May. By then, the budget has been reallocated. The campaign is off. The keyword is paused. And you've thrown away money that was actually working.

This happens constantly. I've seen accounts pause keywords that turned out to be 40% converters. I've seen campaigns killed that generated $500K in pipeline. All because the optimization window was too short.

The worst part is that Smart Bidding makes this worse. The algorithm learns from the data it sees. If it's not seeing the delayed conversions, it's learning from incomplete data. So it's biased toward fast-converting, cheap garbage. It starves the campaigns that generate slow, high-value deals. And you're watching your "efficient" CPL go down while your pipeline dies quietly in the background.

How to actually measure this

The first step is to see the ghost effect in your own account. This requires connecting Google Ads to your CRM and tracking actual closed-won deals, not form fills.

The cleanest way is a simple BigQuery query. Pull your click data from Google Ads. Pull your closed-won opportunities from Salesforce. Join them by the GCLID (or your UTM tracking, if you have it). Then look at the lag between the click date and the close date.

What you'll probably find is a distribution that looks like this:

• 30% convert within 30 days
• 25% convert between 31-60 days
• 20% convert between 61-90 days
• 15% convert between 91-180 days
• 10% convert after 180 days

That's roughly what I see across industrial B2B. The exact numbers vary by product and sales cycle length, but the pattern is consistent. If you're making optimization decisions based on a 30-day window, you're only seeing 30% of your actual conversions. The other 70% are ghosts.

The action: extend your conversion windows and build lagged cohort reporting

You have two moves here.

First, extend your conversion windows in Google Ads from the default 30 days to at least 90 days. Better yet, go to 120 days. Yes, this will make your numbers look worse initially. That's because you're finally seeing the real picture. It's supposed to feel worse. The ghostly, invisible conversions are materializing.

Second, stop looking at day 30 performance data as gospel. Start building cohort reports in BigQuery that track conversion arrival over time. Here's the idea: take all the clicks from January. Don't measure them on January 31st. Measure them on February 28th, then again on March 31st, then April 30th. See when the bulk of conversions actually land.

Then take that information and inform your bid strategy. If you know that most conversions land 75 days out, don't pause campaigns based on day 30 data. You haven't let them cook yet.

Third, if you're using Smart Bidding or any algorithm that needs conversion signal to function, make sure you're feeding it the delayed data. If you're uploading offline conversions, do it programmatically on a schedule that includes full historical lookback. Don't upload last week's deals. Upload last quarter's. Give the algorithm the real picture.

The real number: what this means for your budget

Here's the business impact. You're managing a $100K a month B2B account. Your average closed-won deal is $75,000. Your close rate is 8%.

On day 30, you've generated $40K in pipeline from a $100K spend. That looks like a 2.5X ROAS and a terrible month.

On day 120, that same $100K has generated $180K in pipeline. That's a 4.5X ROAS and a pretty good month.

But if you made optimization decisions on day 30, you probably cut the budget, paused the winners, and shifted money to something that looked better (but probably wasn't). You left $140K on the table.

This is systematic. It's happening in almost every B2B account right now. And the only way to fix it is to stop trusting 30-day windows and start tracking actual business outcomes.

Extend your reporting horizons. Wire up your CRM properly. And stop judging campaigns by the data you have today. Judge them by the data you'll have in 120 days.