Bid strategies for low-volume, high-value maritime queries

Google’s automated bid strategies (Maximise Conversions, Target CPA, Target ROAS, Maximise Conversion Value) are built for accounts with enough conversion data to feed the algorithm. The threshold quoted by Google is around 30 conversions in 30 days. Most maritime accounts produce a fraction of that on direct conversion events. The bidder, faced with sparse data, falls back to broad heuristics that do not match the actual buyer pattern.

The result, in maritime accounts, is automated bidding that either spends nothing (bidder thinks it cannot find conversions and bids nothing) or burns budget chasing low-value soft signals (bidder optimises toward the easy conversions and ignores the long-cycle high-value ones).

This piece is about what to do instead.

Step 1: feed the bidder more signal

Before changing bid strategies, change what the bidder is bidding to.

If your only conversion event is “contact form submitted” and you get four of those a month, no bid strategy will work well. The bidder needs more signal.

Add and tag micro-conversions with realistic values. A brochure download might be worth £80, a case-study view-to-completion might be worth £40, a 75% video view on a product demonstration might be worth £25. These are not direct revenue values; they are the bidder’s input to learn about which clicks behave like buyers.

Add value to the high-intent signals too. A contact form might be worth £600, a qualified RFQ might be worth £3,500, an enterprise enquiry form might be worth £8,000. Set values based on your real average pipeline value times your event-to-opportunity rate.

With twenty to forty conversion events per month across a tagged hierarchy, the bidder has enough signal to start working.

Step 2: pick the right strategy for the account state

There is no universally right strategy. The right one depends on where the account is.

• New account, no conversion history: manual CPC, with deliberate bids set for each keyword group. Run for sixty to ninety days to build conversion history.
• Account with under 30 conversions per month: Maximise Conversions without a target CPA. Lets the bidder use what data it has without forcing constraints it cannot meet.
• Account with 30 to 100 conversions per month and consistent value-per-conversion: Target CPA at a value 10% to 20% above the actual CPA. Constrains the bidder to your zone of acceptable cost.
• Account with 100+ conversions per month and meaningful value variance: Target ROAS or Maximise Conversion Value. The bidder can now do real work because there is enough variance and volume to learn from.

Brand campaigns are the exception: manual CPC works fine on brand because the volumes are stable and the bidding is mostly defensive.

Step 3: protect the account from bidder overreaction

Automated bidders react quickly to short-term data. In low-volume maritime accounts, two weeks of unusually high or unusually low activity can pull the bidder badly off course.

Useful guardrails:

• Max CPC ceilings. Set a maximum CPC bid even on automated bid strategies (Google calls this a “bid ceiling”). Stops the bidder from paying £40 a click on a one-off competitor lookup query.
• Max CPC floors. Less commonly used, but on high-value low-volume queries, a floor stops the bidder from under-investing during quiet periods.
• Conservative target shifts. If you change a Target CPA, change it by 10% to 20% maximum. Larger changes restart the bidder’s learning phase and produce a fortnight of unstable performance.
• Negative-keyword discipline. A clean negative-keyword list is more important under automated bidding than under manual; the bidder will happily spend on tangential queries if the negative list does not block them.

Step 4: use bid adjustments where you have signal

For audiences and segments where you have evidence of higher buyer concentration, bid adjustments still work even on automated strategies (they bias the targets, not the absolute bids).

Useful adjustments in maritime:

• Customer match (existing customers): adjust down 30% to 50% (you do not want to pay full search prices to reach existing customers)
• Customer match (target accounts list): adjust up 20% to 40%
• Geographic adjustments for hub locations (Singapore, Rotterdam, Houston, Piraeus): adjust up where you genuinely serve, adjust down or exclude where you do not
• Device adjustments are mostly noise in B2B maritime; leave at zero
• Hour-of-day and day-of-week adjustments are mostly noise; leave at zero

A note on Performance Max

Performance Max is built for high-volume B2C ecommerce. It does not work well on most maritime accounts. The signal density is too low, the asset variation is too narrow and the lack of query-level visibility means you cannot learn from what is working.

If a generalist agency has set up Performance Max on your maritime account, ask them to justify it. Most accounts under £20,000 monthly spend should turn it off.

What earns the automation

Automated bidding is a powerful tool when the account state genuinely supports it. In low-volume maritime accounts, the work is to build that state up: more conversion signal, accurate values, clean negatives, sensible guardrails. Once that is in place, the automation earns its keep. Without it, automation produces noise that looks like learning.