What does it actually cost a window and door company when an agency’s lead count goes up while its close rate goes down? Most operators cannot answer that question with a number, because the reporting they receive was never built to expose it.
That gap is not an accident of a bad vendor. It is what happens by default when marketing performance and sales performance are measured in two different rooms. The distance between them is where acquisition cost creeps up even as the dashboards look fine.
Why Full-House Replacement Demand Behaves Differently From Generic Home-Services Leads
A full-house window and door replacement is not a same-day repair call. The homeowner is weighing a five-figure decision, comparing at least one other bid, and deciding whether to trust a crew inside their home for several days.
That consideration cycle changes what a lead means. A single contact who never returns a call is not a wasted click, it is a missed contract worth tens of thousands of dollars.
The appointment calendar behind it has real capacity limits too: a sales team can only run so many in-home consultations in a week before the quality of the pitch itself degrades.
Service-area limits compound this. A campaign that generates interest outside the installer’s actual coverage zone produces activity without producing appointments that can close.
The right objective is not more leads. It is more qualified opportunities, meaning inquiries that match the company’s job-size target, sit inside its service area, and arrive with enough information for a sales rep to prepare before the visit.
Scaling contracts, then, is not something an agency does by itself.
It is a coordination problem across:
- demand generation
- the person answering the phone
- the rep closing the estimate
| and a proposal that only addresses the first of those three | has not addressed the business problem at all. |
Which Agency Promises Create the Most Expensive Blind Spots
Vague reporting is not neutral, it is a cost. When a monthly report lists impressions, clicks, and “leads generated” with no definition of what counts as a lead, the number can rise for reasons that have nothing to do with sales, including a change in bid strategy or a shift toward broader match keywords that pull in price shoppers.
Generic home-services playbooks make this worse: tactics tuned for plumbing or HVAC emergency demand do not transfer cleanly to a purchase decision with a longer research phase and a much higher average job value.
A handful of warning signs tend to predict the same downstream problem, an invoice that grows faster than the contract count:
- No lead definition: if “lead” is never defined in writing, the agency and the sales team are grading two different tests.
- No service-area exclusions: campaigns running outside the coverage zone waste both ad spend and a rep’s time qualifying a job that was never winnable.
- Inaccessible accounts: an agency that will not grant account-level visibility into paid search or analytics is asking to be trusted rather than checked.
- Unexamined call quality: nobody listening to what actually happens on the phone means volume can rise while contact rate falls.
- Untestable SEO milestones: “improve visibility” is not a milestone, it cannot be disproven, and it cannot be renegotiated against.
- No exit terms: a contract with no defined way out ties spend to a relationship that has already stopped earning it.
None of this proves a given market behaves any particular way. It proves that a report built without these guardrails cannot tell an operator whether the money is working.
How to Test Agency Expertise Before a Contract Is Signed
The evaluation question underneath every glossy capabilities deck is simple: can this agency show relevant judgment, or only relevant vocabulary? A team that can say “cost per lead” and “quality score” fluently has learned the language of the channel, not necessarily the economics of a full-house replacement business.
The difference only shows up when they are asked for evidence rather than assurance.
A useful way to separate the two is to ask for specific proof against each claim, in this order:
- Replacement-relevant case studies. Credible proof looks like named campaign context, job type, and the reasoning behind channel choices. What stays unproven is any verified revenue or profit outcome tied to that campaign.
- Discovery questions on job value and margin. Credible proof looks like specific questions about average ticket size, install capacity, and target margin, asked before any media plan is proposed. What stays unproven is whether the agency will actually adjust targeting once margins are known.
- A landing-page review. Credible proof looks like a walkthrough of message match between ad copy and page content, with a clear call to action. What stays unproven is the conversion rate on a page the agency has not yet built or tested.
- A competitive search review. Credible proof looks like an honest account of what the search results for relevant terms currently look like, not a promise to dominate them. What stays unproven is any claim about ranking position weeks or months out.
- Paid search manager access and reporting samples. Credible proof looks like a real, redacted account export or dashboard, not a screenshot from a case study deck. What stays unproven is performance identical to what a past client saw in a different market.
A firm that treats the specialization claim itself as evidence is worth a second look. There is a real difference between an agency that says it understands the category and one that publishes methodology a prospect can actually scrutinize.
That kind of detail is what a page like a window and door marketing agency would need to show, not as an endorsement of any single provider, but as an example of the bar a specialization claim should clear.
If the discovery conversation never gets to margin, capacity, or service-area exclusions before a media plan appears, the fluency was cosmetic.
How Paid Search and Local SEO Should Work Together
Paid search and local SEO are not competing for the same job. Paid search exists to capture a homeowner who has already decided to act and is searching right now. Local SEO exists to be visible during the slower research phase that precedes that moment, when the homeowner is comparing companies rather than typing “window replacement quote near me.”
Neither channel excuses a weak qualification process downstream. A perfectly targeted ad that lands on a slow, generic page, or a well-optimized business profile that hands off to an unanswered phone, produces the same result: spend without a signed contract.
How Paid Search Campaigns Should Be Scoped
Campaign scope starts with keyword intent. Terms that signal an active project, “full house window replacement cost” or “replacement windows and doors installer,” belong in a different campaign from broad category terms that draw browsers and researchers.
Negative keywords matter as much as the target list: excluding “repair,” “parts,” “DIY,” and similar terms keeps the budget away from jobs the company does not sell. Service-area exclusions at the geographic targeting level prevent spend from following clicks outside the installer’s actual coverage.
Landing pages need to match the ad’s specific promise, not route every click to a generic homepage. Budget logic should follow job value and sales capacity rather than an arbitrary monthly figure.
A company that can only run twelve consultations a week does not benefit from a fourth campaign generating a hundred more inquiries than it can staff.
A paid search manager’s ongoing job is optimization against qualified appointments, not against clicks or impressions, which means the reporting relationship has to reach past the ad platform and into what the sales calendar actually shows.
What Local SEO Should Make Visible
Local SEO’s job is different: service pages built around specific job types, a maintained business profile, and the local relevance signals that influence whether a company appears when a homeowner is doing early comparison research rather than searching to buy immediately.
Milestones here need to be measurable, more service pages indexed, a business profile with current information and responses to reviews, rather than a vague promise of better visibility.
Context on paid search economics is available, though it should be read as a benchmark and not a guarantee. A 2025 benchmark table published by LocaliQ lists Doors and Windows Sales at a $200.34 Google Search cost per lead, describing it as the second-highest cost-per-lead subcategory in that table.
That figure says something about how competitive and expensive the category can be to bid into. It says nothing about whether a given agency’s leads convert to appointments, or whether a given market matches that benchmark at all.
What Must Be Agreed Before Campaign Work Begins
Campaign work should not start until a small set of operating decisions is on paper.
- The target job profile comes first: what counts as a full-house replacement job worth pursuing, and what disqualifies an inquiry outright, whether that is a single-window repair, a rental property, or a budget range the company does not serve.
- Service areas and their exclusions need to be reviewed against the actual install team’s coverage, not against a broader regional ambition.
- Job-value, margin, and sales-capacity assumptions have to be stated explicitly, since a campaign tuned for a $30,000 average job behaves differently from one tuned for a $12,000 job.
- From there, the website and landing-page conversion paths need a plain review: does the page that receives the click actually ask for the information a rep needs before a consultation?
- Source-to-appointment tracking has to be configured before launch, not added retroactively once someone asks where a contract came from.
- A feedback loop with sales, even an informal one, should be scheduled from the start rather than treated as an afterthought once volume looks off.
The lead-to-contract path itself runs through five decisions:
- Search visibility earns the initial contact.
- The inquiry is qualified or disqualified.
- The qualified appointment is scheduled.
- The estimate is presented.
- The contract is signed or lost.
Each handoff is a decision point, not a formality, and a reporting system that only tracks the first two steps cannot tell anyone where the funnel is actually leaking.
Franchise territory rules and localized strategy matter here only where they apply: a multi-location franchise operator has constraints a single-market company does not, and neither should be treated as a universal requirement.
Which Metrics Show Whether Marketing Is Producing Sales-Ready Demand
Platform metrics describe what an ad platform did. They do not describe what a sales team received. Click-through rate, impressions, and even cost per lead in isolation say nothing about whether the person on the other end of that lead was ever going to buy a full-house replacement.
Assessing return on ad spend requires a reporting hierarchy that follows the money past the point where the agency’s job technically ends.
That hierarchy has a specific shape:
- Spend and source: what was spent, and through which campaign or channel it arrived.
- Inquiries: the raw count of contacts, before any qualification.
- Contact rate: how many inquiries were actually reached by phone or otherwise.
- Qualified appointments: inquiries that matched the target job profile and were scheduled.
- Estimates presented: appointments that converted into an actual quote.
- Contracts signed, where available: the number that ties spend back to revenue.
- Attribution limitations, documented: an honest note on what the tracking cannot see, rather than silence where the gap sits.
A reporting cadence should be agreed in writing, along with who owns the ad accounts and analytics access, and a shared lead definition that both the agency and the sales team accept before the first invoice arrives.
Call recording or a qualification review, where legally permitted, closes the loop between what marketing generated and what sales actually worked. Sales-team feedback, delivered on a schedule, is the mechanism that lets targeting improve over time, and a partner who resists that feedback loop is choosing not to improve.
What the Final Agency Request Should Ask For
Every criterion above converts into a single comparable request that can be sent to more than one prospective partner at once. That request should ask for evidence of category fit, not a claim of it:
- relevant case studies or portfolio context
- the specific channels in scope
- who owns the ad accounts and website assets once the engagement ends
- how lead quality gets reviewed and by whom
- what local-SEO methodology looks like in practice
- what the reporting cadence covers
- how budget gets allocated across channels and why
- what the exit terms are if the partnership does not perform
The uncertainty that opened this comparison was never really about lead volume.
It was about the unasked cost sitting behind a marketing promise that sounds complete but cannot be inspected. The test that resolves it is whether a partner will let its assumptions and its measurement be checked, not whether its pitch sounds fluent.
An agency that welcomes that scrutiny has nothing to lose from it. One that resists it has already answered the question.
A request worth sending should read something like this:
Show your discovery process for job value, margin, and service-area fit before any media plan. State your paid-search and local-SEO scope, with campaign structure and negative-keyword logic. Define what counts as a qualified lead, and describe the feedback loop with our sales team. Confirm who owns ad accounts, analytics, and website assets during and after the engagement. State the reporting cadence and the exit terms if performance does not hold.