The most expensive Google Ads campaign mistakes often happen before the first useful enquiry arrives. A poorly defined goal, an untested conversion and a landing page that does not match the offer can reinforce one another. Buying more traffic will not tell you which of those problems needs fixing. Start by checking the complete journey from a search to a business outcome.
This guide covers ten mistakes for business owners and marketing teams preparing or reviewing a search campaign. It is a diagnostic framework, not a promise that one bidding strategy or account structure will suit every advertiser. The numerical examples are hypothetical. They are not WeAreMedia client results, forecasts or industry benchmarks.
1. Buying clicks without agreeing what success means
“Increase website traffic” and “generate suitable requests for a quotation” describe different jobs. A campaign can deliver inexpensive visits while producing few enquiries that a sales team can use. Before opening campaign settings, define the customer, the offer, the area you can serve and the action that would represent progress.
For a service business, a qualified enquiry might require a relevant need, a serviceable location and a reachable contact. For an online store, an order has to be understood alongside margin, fulfilment, cancellations and returns. Agree these definitions with the people who handle the actual sales. The advertising report and the sales team should not use the same word for different outcomes.
Write a short objective that someone outside marketing can understand. For example: “Generate requests for the commercial installation service from businesses within our service area.” Then identify the evidence needed to judge the requests. A form submission is an observable step; suitability and a completed sale require further information.
Traffic can be a legitimate objective for a campaign designed to build awareness or introduce a resource. The mistake is evaluating that campaign as though it had been designed to produce immediate sales. Keep its role explicit and allocate a budget the business can justify for that role.
2. Assuming a tracking tag proves conversions are correct
A tag being present does not establish that the intended business action is counted correctly. Opening a form, clicking its submit button and completing a successful submission are different events. A failed form can still produce a button click. A messaging link click does not establish that a message was sent or that a suitable enquiry followed.
In Google Ads, primary actions generally inform bidding when their standard goal is used by the campaign. Secondary actions normally provide observation, with an exception for actions included in custom goals. Check the campaign's actual goal selection as well as the action label. Google's primary and secondary conversion documentation explains the distinction.
Test a successful transaction, a failed attempt and a repeated visit to the confirmation page. Use a phone as well as a desktop. Check whether the same purchase is being treated as a primary result through more than one measurement source. Two records describing the same business outcome should not automatically be added together as two separate successes.
Keep a simple measurement specification: event meaning, trigger, source, value, counting rule and test result. Include the effect of consent and any gaps you cannot measure. If the implementation needs work, conversion tracking and analytics support should address the whole customer journey rather than merely adding another tag.
3. Treating keyword relevance as buying intent
A search can contain your main service term without expressing demand for that service. Someone researching a course, looking for a job or comparing suppliers may use overlapping words. Putting all those needs into one ad group makes it difficult to write a useful message or choose a suitable destination.
Matching also goes beyond literal spelling. Even exact match can involve meaning and intent; it is not a guarantee that an ad will appear only for one identical sequence of characters. Review Google's explanation of keyword matching before treating punctuation around a keyword as a complete control system.
Separate the needs you want to serve: buying, evaluating alternatives and learning. Decide which deserve paid traffic under the current objective. Then assess match types alongside measurement reliability, bidding, budget and your capacity to review the resulting searches. Broad match is not automatically wrong, and exact match does not remove the need for oversight.
Consider a manufacturer that supplies bespoke packaging in large quantities. A person looking for one ready-made gift box may be a poor fit despite the shared vocabulary. Stating a minimum order in the ad might reduce clicks while improving the usefulness of the enquiries. Relevance includes commercial conditions, not just a product category.
4. Ignoring search terms or excluding too much
Keywords are advertiser inputs; search terms reflect what people searched for. The Google Ads search terms report helps you examine the visible queries associated with ad delivery. Some low-volume queries are omitted for privacy, so it is not a complete list of every search.
Look for clearly unsuitable needs, useful enquiries and searches that deserve a better answer on the destination page. Do not automatically exclude a relevant term after a handful of clicks without a conversion. A small sample and a clear mismatch are different findings. Your acceptable test cost should help determine when there is enough evidence to act.
Negative keywords need context too. If one part of the business sells training, excluding “training” throughout the account may remove customers from that part of the offer. Review the negative's match type and scope before applying it. An account-wide exclusion should have an account-wide reason.
Record the term, the problem, the action and the evidence. “Expensive” is insufficient if a higher-cost query generates customers with greater value. Where possible, connect the review to sales feedback instead of sorting only by cost per click. Also check whether an apparent mismatch actually exposes a useful question your existing page fails to answer.
5. Choosing locations and languages without an operational reason
On-site work, remote consulting, local delivery and international sales require different decisions. Targeting the owner's city by habit can miss valid demand; targeting an entire country can produce requests the business cannot fulfil. Write down the service boundaries before choosing the advertising boundaries.
Location options distinguish presence from interest in a location, and location signals are not perfectly accurate. Review Google's advanced location options against the actual service model. A broader interest setting and a presence-focused setting answer different needs.
For example, a repair team that only visits nearby premises may have little use for distant demand. An event supplier, however, could receive a valid request from someone abroad arranging an event locally. A universal rule to target only people physically present would overlook that second situation. Review the customer journey instead of adopting a setting as a slogan.
Language choices must also match the ad, destination and follow-up process. An English ad leading to a form whose terms are available only in another language creates uncertainty. Confirm that the business can explain the proposal and handle enquiries in the language it advertises. More requests are not helpful if nobody can serve them reliably.
6. Misreading the budget and spreading it too thinly
An average daily budget does not mean identical spending every day. For most campaigns, daily spending can reach twice that average; the general monthly charging limit for an unchanged budget is the daily average multiplied by 30.4. Review the conditions for your campaign and budget changes in Google's spend management guide.
As a hypothetical example, an unchanged £50 daily average across a full month gives a general monthly calculation of £1,520. That is a media calculation, not a complete project price. Taxes, agency work and creative production may be separate. Exceptions, including some payment models, mean you should check the actual billing rules rather than assuming the same daily ceiling applies everywhere.
The second problem is splitting a limited budget across too many campaigns, offers and regions. When each part receives very little traffic, one order can change the entire interpretation. Separate campaigns when their objective, economics or operational constraints justify it. More folders in an account do not necessarily produce clearer decisions.
Agree a monthly media allowance, a test allowance and the conditions for reviewing spend. Identify who can approve a change and who watches delivery. A modest budget still requires a reliable outcome definition. Spending slowly against the wrong signal is not a sound testing strategy.
7. Setting bidding targets from wishes instead of evidence
The acquisition cost a business would like to achieve may differ from what its current offer and customer journey can support. Entering an ambitious target does not repair a weak proposition, a confusing page or unreliable measurement. Choose the approach according to the campaign's purpose and the quality of its available signals.
Recent results can also be incomplete. Someone may click today and request a proposal next week. Cost appears before the resulting opportunity. Google's advice on measuring Smart Bidding performance recommends considering conversion cycles and allowing results to settle after significant changes.
Changing targets, budgets, goals and landing pages in quick succession makes interpretation harder. Fix a clear malfunction first, then test a specific hypothesis. Waiting for performance to stabilise is appropriate for a working system; it is not a reason to keep paying for traffic to a broken form.
For each change, record the previous state, the expected effect and the review conditions. Avoid a universal rule that every account can be judged after exactly seven days. A long sales cycle and an inexpensive immediate purchase need different observation periods. A few conversions cannot support the same confidence as a substantial, consistent sample.
8. Breaking the promise between the ad and the landing page
An ad for a specific offer should not leave the visitor searching a general homepage to find it. The destination needs to confirm that the visitor is in the right place, explain the offer and make the next step understandable. The headline, conditions and action should all support the same decision.
If the ad offers a free assessment, explain what information is required and what the person will receive. For a product, make availability, delivery and material costs clear. Hiding a condition may increase initial enquiries while creating disappointment later. A useful landing page helps unsuitable visitors recognise the mismatch as well as helping suitable ones proceed.
Test the complete task on a phone. Find the relevant information, fill out the form, correct an error and finish the action. Check whether the keyboard obscures a field, whether the error message is helpful and whether the contact links work. A desktop screenshot or a speed score cannot replace that exercise.
The principles in our guide to turning social media interest into website conversions also apply here: preserve the promise made before the click and remove avoidable uncertainty afterwards. You do not need a different page for every keyword. You need a clear destination for each genuinely different offer or decision.
9. Reporting every form as an equally valuable lead
Twenty submissions can contain very different outcomes: suitable requests, wrong-service enquiries, unreachable contacts and duplicates. If the advertising team only sees the total, it may celebrate a lower cost per form while the sales team sees fewer usable opportunities. Connect the two views through a shared definition and feedback process.
Start with a minimal record of date, source, service requested, qualification status and sales outcome. Collect only the information the business needs and limit access to appropriate staff. Distinguish “not contacted,” “unsuitable,” “proposal sent” and “won.” Those statuses describe different problems and should not be collapsed into one conversion total.
Hypothetical comparison: Campaign A spends £1,000 and generates twenty submissions, two of which are qualified. Campaign B spends the same amount and generates ten submissions, five of which are qualified. Cost per submission is £50 versus £100, while cost per qualified enquiry is £500 versus £200. The cheaper form is not automatically the better business result. Sales and contribution still need further assessment.
For ecommerce, the same issue appears when revenue looks strong but the orders leave little contribution. Our ROAS and break-even guide connects reported returns with product economics. In service businesses, also track response handling: losing a suitable enquiry because nobody follows up is a different problem from attracting the wrong visitor.
10. Changing everything and then claiming to know what worked
A new ad, a different destination, revised targeting and a new conversion goal introduced together create a difficult comparison. Even if performance improves, you cannot confidently attribute the change to one intervention. Account management should produce understandable decisions, not a constant stream of adjustments.
Put the reporting period, campaign objective, conversion definition and significant changes at the top of the review. Separate searches for the existing brand from efforts to reach people discovering the business. Note stock availability, price changes, seasonality and sales capacity where they affect interpretation. A single click-through-rate threshold cannot fairly judge every intent, device and position.
Prioritise faults in a practical order: broken measurement or customer actions, clearly unsuitable traffic, then offer and page alignment. Give each action an owner and a review point. If an agency manages the account, the business should still be able to access its own reports and understand the reasoning behind changes.
When considering Google Ads management support, ask which finding motivates the proposed action and which business outcome it is intended to improve. A higher score or more clicks may be useful diagnostic information, but neither answers that question on its own.
A short launch and review checklist
- Is the intended customer and offer clear, including situations the business cannot serve?
- Has the main conversion been tested for successful, failed and repeated actions?
- Do search intent, advertising copy and destination answer the same need?
- Can the business handle the chosen locations and languages?
- Are media spend and the total project cost understood separately?
- Can a visitor complete the intended action on a phone?
- Who will assess the quality of enquiries or the economics of orders?
- What evidence will justify the first change?
If most answers are unclear, launching more campaigns will add activity without resolving the uncertainty. For an account already running, you do not necessarily need to rebuild everything. Validate the measurement, identify the most consequential loss and make one documented correction. Keep the existing history so the next review has something meaningful to compare.
What if there are clicks but no conversions?
Check the conversion journey and reporting delay first. Then examine search intent, destination and offer conditions. An obvious malfunction can justify stopping affected spend while it is repaired. A small number of visits alone, however, does not establish that a relevant campaign cannot work. Judge the available evidence against the test budget agreed in advance.
These checks do not guarantee cheaper leads or better rankings. Competition, demand and the offer still matter. Their purpose is to reduce preventable measurement errors, expose unsuitable spending and make the next decision easier to explain. Revisit the relevant checks when the product range, service area, form or sales process changes: an earlier measurement definition may no longer describe the business outcome you now need.