Google Ads When the Sale Takes Six Months

Google Ads for long sales cycles needs an earlier truth

A six-month sale should not usually be the conversion Google Ads bids towards. Closed-won is financially honest, but it reaches the bidder after the auctions it needed to influence. For Google Ads for long sales cycles, the fix is a nearer, honest signal—not a cleverer bid strategy.

Finance should still judge the investment against closed-won contribution. Bidding needs a qualified event that occurs frequently enough and quickly enough to influence the next auction.

Confusing those jobs creates a predictable failure. Revenue columns remain empty while deals mature, so the manager cycles through Maximise Conversions, target CPA and target ROAS. Every switch answers a timing problem with a control change.

The surrounding commercial and account decisions belong in our full B2B Google Ads strategy guide. The delayed-feedback problem needs its own design because a campaign can be commercially productive while looking barren inside Google Ads.

The correct question is not, “Which bid strategy is best?” It is, “Which truthful event can the chosen strategy observe in time?”

Delayed revenue is a reporting truth, not always a bidding signal.

The signal-choice tree permits the bid strategy

Start with two fields for each campaign: eligible conversions per complete month and median time from ad click to that conversion. Calculate both from mature cohorts, excluding recent clicks that have not had time to progress.

Conversion volume per month and time-to-conversion determine which strategy is permitted. They are gates, not preferences.

Use the following four-branch decision tree:

1. At least 30 closed-won conversions per month, with median lag no longer than 30 days: closed-won can remain the Primary conversion. Maximise conversion value is permitted; target ROAS becomes permissible after two consecutive 30-day periods each contain at least 30 outcomes.
2. Closed-won fails either test, but a qualified CRM stage produces at least 30 conversions monthly with median lag no longer than 30 days: make that stage Primary. Use Maximise Conversions, or Maximise conversion value when the stage values have an economic basis.
3. The nearest honest stage produces 15–29 conversions monthly with median lag no longer than 45 days: consolidate like-for-like campaigns and use Maximise Conversions without a target. Reconsider target CPA only after two consecutive months reach 30 conversions.
4. Every honest stage produces fewer than 15 conversions monthly or has median lag above 45 days: target CPA and target ROAS are not permitted. Use Manual CPC or Maximise Clicks with a CPC ceiling while fixing measurement or consolidating demand.

These are operating thresholds, not published Google Ads guarantees. The honest limit is that volume alone cannot describe signal quality. Seasonality, duplicate events and one oversized opportunity can make 30 observations less useful than the count suggests.

Our claim is falsifiable: below 15 signals monthly with median lag above 45 days, changing only the bid strategy will not reduce mature 90-day cost-per-qualified-opportunity variance; a controlled holdout that does would prove us wrong.

Terms such as Primary conversions, target CPA and conversion value are defined in our plain-English Google Ads glossary.

Signal eligibility comes before strategy selection.

A stage-value ladder for long sales cycle Google Ads

A raw form submission is fast but rarely honest. Closed-won is honest but usually too slow. The useful midpoint is a CRM event whose definition cannot be changed after somebody sees campaign performance.

Use economic values rather than arbitrary scores such as 10, 50 and 100. The following four-rung illustrative ladder assumes £20,000 expected contribution from an eventual deal:

1. Accepted ICP enquiry: assumed win probability is 5%. Upload £1,000.

5% × £20,000 = £1,000 cumulative expected value

2. Discovery meeting attended: assumed win probability rises to 10%. Upload another £1,000.

(10% × £20,000) − £1,000 already uploaded = £1,000 incremental value

3. Sales-qualified opportunity created: assumed win probability rises to 25%. Upload another £3,000.

(25% × £20,000) − £2,000 already uploaded = £3,000 incremental value

4. Closed-won: actual contribution is £20,000. Upload the remaining £15,000.

£20,000 − £5,000 already uploaded = £15,000 incremental value

Incremental values prevent one journey being counted four times at full value. Upload the four milestones as separate conversion actions, then place the earliest rung that passes the decision tree in the Primary bidding goal. Keep unqualified form submissions outside that goal.

Closed-won-only bidding starves the algorithm. In an illustrative comparison, 30 accepted enquiries per month provide a recurring signal, while two closed deals per quarter provide a delayed correction. The earlier event wins as the bidding input; the later event remains the financial verdict.

Defining an accepted lead is a separate discipline covered in our guide to designing B2B lead generation around quality. Stage values should use expected contribution rather than headline contract revenue; our method for calculating commercially honest ROAS explains that distinction.

If the mature win rate changes by more than 5 percentage points, or average contribution changes by more than 20%, recalculate every rung.

Expected value gives the bidder speed without inventing success.

Set attribution against the real conversion lag

Pull two lag distributions from complete CRM cohorts: click to the chosen Primary event, and click to closed-won. Median lag governs bidding eligibility; the 90th percentile governs how much reporting time the account needs.

Set the Google Ads click-through conversion window no shorter than the 90th-percentile lag for the Primary event. If that lag exceeds the longest available platform setting, select the longest setting and use CRM cohort reporting for the remainder. The interface figure is then incomplete by construction.

Maintain two reporting views. The operational view groups recent activity by conversion event date. The mature cohort view groups eventual revenue by the original click month. Mixing them produces a figure that is neither timely enough for optimisation nor mature enough for finance.

A 30-day view and a 180-day cohort view can report radically different ROAS for the same campaign without any change in commercial performance. Extending the window changes recognition, not economics.

This separation is central to our Google Ads campaign management for B2B firms, because interface ROAS cannot adjudicate outcomes it has not yet received.

A short window makes a long cycle look unprofitable.

When account metrics and sales outcomes diverge, we trace the cause from search term to CRM — book a call

A worked UK B2B example with two ROAS answers

Take an illustrative UK industrial consultancy selling high-value compliance projects. Its five labelled inputs are:

• Monthly media spend: £8,000.
• Clicks: 400 at an average CPC of £20.
• Mid-funnel conversions: 20 accepted ICP enquiries, arriving within a median of 18 days and valued at £1,000 each.
• Closed deals: two from the click cohort, closing on days 145 and 168.
• Contribution per deal: £30,000 after delivery costs.

The arithmetic is reproducible:

£8,000 monthly spend ÷ £20 CPC = 400 clicks

20 accepted enquiries ÷ 400 clicks = 5% accepted-enquiry rate

20 accepted enquiries × £1,000 stage value = £20,000 expected value

£20,000 expected value ÷ £8,000 spend = 2.5x 30-day signal ROAS

£0 closed-won contribution within 30 days ÷ £8,000 spend = 0.0x reported ROAS

(2 deals × £30,000 contribution) ÷ £8,000 spend = 7.5x 180-day cohort ROAS

The same campaign therefore reports 0.0x closed-won ROAS at 30 days versus 7.5x in the mature 180-day cohort. The 2.5x stage-value figure is an early expected-value signal, not booked return.

Killing the campaign after 30 days would discard a cohort that later produces £60,000 contribution. Calling it a success after 30 days would also be premature because the assumed stage value still needs validation.

If three mature cohorts show an accepted-enquiry win rate below 3% versus the assumed 5%, reduce the stage value before changing bids.

Mature cohorts separate weak demand from delayed revenue.

What does not work in delayed-conversion campaigns

Delayed feedback attracts visible activity that leaves the underlying data unchanged. Four supposed fixes fail repeatedly.

1. Switching bid strategies every fortnight: moving from Maximise Conversions to target CPA or target ROAS does not manufacture missing observations. It changes how the same incomplete evidence is used.
2. Judging the latest 30 days when median click-to-sale lag is 150 days: the recent cohort has barely started maturing. Compare complete 180-day cohorts instead of declaring the newest month unprofitable.
3. Promoting page views, 60-second sessions or PDF downloads to Primary conversions: these events create volume by lowering the definition of success. The bidder then learns to buy inexpensive curiosity. Poor query intent requires a separate B2B keyword strategy built around commercial intent, not softer conversion goals.
4. Allowing recommendations to rewrite the account automatically: Google Ads optimises against the goals and data available inside Google Ads, not the board’s full commercial context. Open Recommendations → Auto-apply and disable automatically applied changes across five categories: Bidding, Budgets, Keywords and targeting, Ads and assets, and Measurement. The sole arguable exception is Repairs, restricted to restoring an existing approved ad or destination.

No bid strategy can compensate for an absent learning signal.

FAQ

Four operational questions determine whether delayed conversion data survives contact with the CRM.

How often should CRM-to-Google Ads mappings be checked?

Audit them every 30 days using a random sample of at least 10 journeys. If more than one journey has the wrong stage, value or click identifier, pause value-based bidding until the mapping is corrected.

Should tracked phone calls count as Primary conversions?

Only when the caller is matched to the originating ad and sales accepts the enquiry within seven days. Unreviewed calls remain Secondary conversions regardless of call duration.

How should cancelled contracts affect conversion value?

When more than 5% of won deals cancel within 90 days, use expected net contribution rather than booked contract value. Conversion adjustments should remove or reduce reversed value.

How should international deal values be handled in a UK account?

Normalise every uploaded value to GBP using one finance-approved monthly exchange rate. If a currency moves by more than 5% during the month, finance should approve an interim rate change.

CRM discipline determines whether delayed conversion data remains usable.

Summary

Five operating rules govern delayed-conversion campaigns:

• Use closed-won for bidding only at 30 or more monthly outcomes and lag no longer than 30 days.
• At 15–29 timely signals, consolidate comparable campaigns and avoid target CPA.
• Below 15 signals or above 45-day median lag, prohibit target CPA and target ROAS.
• Set attribution to the 90th-percentile lag and use CRM cohorts beyond the platform’s reach.
• Disable auto-applied changes to bidding, budgets, targeting, creative and measurement.

Honest signals beat delayed outcomes.

Actualyse runs precision B2B Google Ads programmes optimised for lead quality, not click volume. Book a call to talk through where yours stands.