Home / Paid Media / Google Ads
Paid Media
Google Ads managed against pipeline, not form fills
Search is the only channel where someone has already told you what they want. Most accounts waste that by optimising toward the cheapest conversion event available. We rewire the signal first, then rebuild the account around it.
Overview
The account is usually fine. The thing it is aiming at is wrong
When we audit a Google Ads account that is underperforming, the problem is rarely bid strategy or ad copy. It is that “conversion” means a contact-form submission, Smart Bidding has spent six months learning which people submit contact forms, and it has become excellent at finding them — students, vendors, recruiters, and buyers far too early to be worth a sales call. Cost per conversion looks respectable in the interface and nothing lands in the pipeline. The fix is unglamorous: define the event that corresponds to a real sales outcome, get it flowing back into Google from your CRM, let the bidding relearn, and restructure the account so the restructure is worth the relearning period.
What’s included
- Full account, structure and conversion-tracking audit before anything is changed
- Offline conversion import or enhanced conversions wired from your CRM
- Search campaign rebuild with intent-segmented ad groups and negative hygiene
- Performance Max and shopping where the catalogue or asset base supports it
- Ad copy and extension testing against pipeline-stage data
- Monthly written reporting separating brand from non-brand spend
Part of our Paid Media engagement · from transparent pricing.
Structure
What we change in the account, and in what order
Order matters more than any individual change, because every structural change resets the learning the bidding has accumulated. Do them in the wrong sequence and you pay the relearning cost three times instead of once.
1. Measurement
Conversion actions redefined against real sales stages, offline imports connected, consent and server-side tagging checked. Nothing else happens until this is right.
2. Waste removal
Search-term review, negative keyword lists, placement and audience exclusions, and the brand-term separation that makes reporting honest.
3. Intent segmentation
Ad groups split by what the searcher actually wants — comparison, pricing, problem, vendor replacement — because each deserves a different page and a different ask.
4. Landing alignment
Each campaign pointed at a page that answers its specific query, rather than all of them at the homepage or one generic demo form.
5. Bidding
Strategy chosen for the data volume you actually have. Target CPA on forty conversions a month is a random number generator, and we will say so.
6. Expansion
Only once the core is converting: Performance Max, shopping, broader match types, and the adjacent channels that the search data has proven demand for.
If an account is in genuinely bad shape, a parallel rebuild beside the existing campaigns is safer than editing in place — you keep the current lead flow while the new structure gathers data.
Keywords
Intent segments, not a keyword list
A flat keyword list sorted by volume is the least useful artefact in paid search. What a searcher wants differs enormously between queries of similar volume, and that difference decides the ad, the page and the offer — not the match type.
Problem-aware queries
Someone describing a symptom rather than a solution. High volume, low immediate intent, best served by a page that diagnoses before it sells.
Category queries
The buyer knows what kind of thing they need. This is where comparison content and clear scoping beat a demo form.
Vendor and alternative queries
Searches for a competitor or ‘alternative to X’. Small volume, exceptional intent, and the queries most accounts are too polite to bid on properly.
Pricing queries
The highest-intent commercial signal there is, and the one most B2B advertisers send to a page that refuses to state a price.
Implementation queries
‘How to migrate’, ‘how to integrate’ — often late-stage buyers checking feasibility, frequently misread as informational and excluded.
Brand queries
Defensive, reported separately, and never counted as incremental. If brand spend is propping up the account's numbers, you should be able to see that.
Automation
Performance Max and Smart Bidding, with the guardrails on
Google's automated campaign types are not the problem people make them out to be. They are extremely effective at finding more of whatever you tell them is good, which is precisely why they are dangerous in an account with a weak conversion signal. The work is in constraining them, not avoiding them.
Signal quality first
Automation inherits your conversion definition. Feed it qualified-opportunity events and it compounds; feed it form fills and it industrialises the wrong audience.
Brand exclusion
Account-level brand exclusions so Performance Max cannot quietly harvest your own brand searches and report them as new demand.
Asset group separation
Separate asset groups per product line or segment, so you can read performance and swap creative without blowing up the whole campaign.
Search theme discipline
Search themes and audience signals used as direction rather than as a wish list, with negative keyword lists applied where the campaign type allows.
Manual structure retained
Enough conventional search structure to keep visibility into which queries and segments produce money, instead of one black box per account.
Known limits stated
Where reporting genuinely does not expose placement or query detail, we say so rather than inventing a confident explanation for a movement we cannot see.
The work
What a Google Ads engagement actually contains
A fixed monthly fee, published on the pricing page, and this shape of work behind it.
The audit is free and useful on its own: if the answer is that your current agency is doing fine and the problem is the offer, that is what the audit will say.
Honest limits
When Google Ads is the wrong place to spend the money
Search captures demand; it cannot manufacture it. There are categories where this channel will not work at the budget available, and knowing that in week one is worth more than a quarter of optimisation.
No search volume
A genuinely new category has no queries to bid on. The budget belongs in demand creation and content until the searches exist.
Unit economics too thin
When a click costs a meaningful fraction of the margin on a sale, no bidding strategy rescues it. The offer or the price has to change first.
Tiny addressable market
A few hundred target accounts means search volume too small to optimise against. Account-based targeting on LinkedIn fits better.
No CRM stage data
Without reliable deal stages there is nothing to optimise toward but form fills, and we are straight back to the original problem.
Broken sales follow-up
If leads wait days for a reply, more leads is the wrong purchase. Lifecycle and automation work comes first — that is a different engagement.
Seasonal cliff edges
Some categories have weeks that matter and months that do not. Spreading budget evenly across the year is a reporting convenience, not a strategy.
In each of these cases we will point at the thing that actually needs fixing, including when that thing is not something we sell.
More Paid Media
The rest of the paid programme
Paid media overview
How the channels fit together and which ones your model actually needs.
LinkedIn Ads
Targeting by title, company and seniority when search volume is too thin.
Paid social
Demand creation and retargeting at CPMs search cannot match.
Landing pages & CRO
The page the ad lands on, built per intent segment and then tested.
SEO
Owning the queries you are currently renting by the click.
CRM
The stage data that makes any of this measurable in the first place.
Questions
Google Ads questions, answered properly
Do you charge a percentage of ad spend?
No. A percentage of spend pays us more for spending more, which is a direct conflict with the job. Our fee is flat, scoped against the work involved — account complexity, number of campaign types, creative and landing-page volume — and published on the pricing page rather than quoted privately.
How long before the account improves?
Expect the waste removal to show within two to four weeks, because cutting spend on queries that never convert is immediate. The bidding changes need a relearning period, typically two to three weeks per significant structural change, and pipeline-stage results cannot appear faster than your sales cycle. If your average cycle is ninety days, nobody can show you closed-won attribution in month one — anyone promising that is reporting form fills.
What budget do we need for this to be worth doing?
Enough that the account produces statistically meaningful data in weeks rather than quarters, which depends entirely on your cost per click. Industrial niche terms and consumer finance terms differ by more than tenfold, so a universal number would be a made-up one. We check your category's live auction costs during the free audit and tell you the floor before you commit to anything.
Can you fix an account without rebuilding it?
Often, yes — and when that is true we say so, because a rebuild costs you a relearning period. Measurement fixes, negative keywords, brand separation and landing-page alignment can be done in place. We recommend a rebuild only when the structure actively prevents reading the data, and then usually in parallel so your existing lead flow continues.
Who owns the account and the data?
You do. Campaigns, conversion actions, tags, audiences, feeds and the reporting build live in your own Google and analytics accounts under your billing, and we operate as a user with access that you can revoke. Nothing sits inside an agency container that you would have to negotiate for on the way out.
Will you bid on competitor brand names?
Where it is legal in your jurisdiction and sensible for your category, yes — those queries carry some of the strongest intent available, and they are cheap because competitors are usually squeamish about them. We do not use a competitor's trademark in ad text, which is the line that causes trouble, and we will tell you if your category is one where this invites retaliation that costs more than it earns.
How do you report, and how often?
A monthly written review, plus a live dashboard you can open any time. Brand and non-brand are always separated, spend is attributed to pipeline stages where the data supports it, and anything we cannot measure is labelled as such rather than estimated into a chart. The review always ends with what we are changing next and what we need from your side.
Do you work alongside an in-house team?
Frequently. Common splits are that we own measurement and structure while the in-house team owns creative, or that we run non-brand while they keep brand. What we need is clarity about who changes what, because two people adjusting bid strategies on the same campaign is worse than either doing it alone.
What if the audit says we should spend less?
Then it says that. We have told prospects to cut budget, to pause a channel, and to fix their sales follow-up before buying more leads. An audit that always concludes ‘you need a bigger programme, which we happen to sell’ is marketing material, not an audit.
How does Google Ads fit with the SEO and GEO work?
Paid search is the fastest keyword research instrument available: within weeks it tells you which queries produce opportunities rather than merely traffic, and which messages convert on each of them. That evidence decides which organic pages are worth building and which questions the pages we want AI assistants to cite should answer. Running the two programmes separately is how companies end up paying by the click for queries they already rank for organically — so we report them in one view and look for exactly that overlap.
Do you work with Microsoft Ads as well?
Where the data supports it. Microsoft Ads carries a meaningfully different audience in some B2B categories, often at lower cost per click, and campaigns can be imported rather than rebuilt from scratch. It is rarely a priority ahead of fixing the Google account, but once the structure and measurement are right it is a cheap incremental test, and we will include it in the plan when your category's search volume there justifies the management overhead.
Ready when you are
Let’s find your next 30% of growth.
A free audit across SEO, GEO, CRM & automation — no strings, no ‘contact for pricing’.