5 Questions to Ask Your Marketing Agency Before Spending Another Dollar

TL;DR: Most marketing agencies can’t answer five critical questions about your ad spend: who visited your website without converting, which visitors they can identify by name, whether remarketing is personalized or generic, which ad dollars drove actual showroom visits, and how fast they pivot when campaigns fail. Up to 98% of website visitors remain anonymous, yet technology exists to identify 25-50% of them. The agencies that can’t answer these questions are optimizing for signed contracts, not successful campaigns. These are the critical questions to ask your marketing agency.

5 Questions To Ask Your Marketing Agency?

If you want to evaluate whether your marketing agency is truly accountable, ask these five questions:

Can you identify anonymous website visitors?
Are campaigns personalized based on vehicle interest?
Can you attribute ad spend to showroom visits?
How quickly do you pivot failing campaigns?
Have you tested vendor overlap and budget cannibalization?

Core Questions Every Marketing Agency Must Answer

Can your agency identify anonymous website visitors by name and provide verified contact information? (Technology exists to identify 25-50% of visitors versus the 2-3% who fill out forms)
Are you running personalized omnichannel campaigns based on specific vehicle interest, or just generic retargeting?
Can your agency draw direct attribution lines from ad spend to showroom visits and phone calls, not just “assisted conversions”?
Does your agency have a pre-agreed decision framework to pivot failing campaigns within 14-21 days, or do they stall with “wait for more data”?
Have you tested your third-party vendor stack (CarGurus, Autotrader, Cars.com) by pausing one at a time for 30 days to eliminate overlap and cannibalization?

I was sitting in a monthly review with a car dealership GM who’d been spending about $50K a month on digital advertising. Pretty standard meeting until the owner asked what seemed like a simple question:

“Who are these people visiting my site? Show me the ones who looked at F-150s yesterday.”

The agency rep started talking about “aggregate data” and “industry-standard analytics.” Basically, a lot of words that meant “no, we can’t tell you that.”

This wasn’t some small shop. This was a well-known agency.

That’s when it hit me. We’re talking about someone writing checks for $600K a year, and they can’t even get a list of people who spent time on their most expensive inventory pages. The agency could tell you how many people visited, what pages they looked at, how long they stayed. But they couldn’t tell you WHO.

Not a single name. Not one email address they could follow up with.

It was like paying someone to count people walking past your store but never actually inviting anyone inside.

Why Most Marketing Agencies Cannot Identify Website Visitors

Here’s the uncomfortable truth: up to 98% of website visitors leave without identifying themselves. You’re running ads, driving traffic, watching colorful charts trend upward in monthly reports—and you have no idea who 98% of those people are.

Most agencies have trained dealerships to never ask the question.

They’ve normalized dashboards that show aggregate metrics instead of actual contacts. They’ve made “we track everything” sound like an answer when it’s really just a dodge.

But there’s enough technology available right now to identify 25-50% of anonymous website visitors.

The technology exists. Many agencies just aren’t agile enough to adapt to it quickly. This directly impacts your ability to maximize your marketing budget.

Bottom line: Technology to identify 25-50% of anonymous visitors exists today, but most agencies haven’t adopted it because they lack agility. This gap costs you leads and budget efficiency.

The 5 Questions to Ask Your Marketing Agency

If you’re spending serious money on marketing, your agency should be able to answer these questions with real numbers and specifics. Not vague promises. Not “we’re working on that.” Actual data.

1. Can you tell me exactly how many people visited our website last month and left without contacting us?

Most agencies will give you total traffic numbers. That’s not what I’m asking.

I want to know: of the people who visited, how many left without filling out a form? Because here’s what matters—only 2-3% of website visitors actually fill out a form on a dealer’s website.

By the time a website shopper fills out a form and reveals their contact information, 70% of their research and shopping has been completed. They’ve already done most of their homework. They’ve compared models. They’ve checked prices. They’ve narrowed their choices.

You’re missing the entire silent research phase, where you could actually influence their decision.

Discover how SMART Leads AI can enhance your lead generation efforts by capturing and identifying anonymous visitors on your website.

2. Of those people, how many can you identify by name and give me actual contact info for?

This is where most agencies stumble.

When you can identify anonymous visitors during that research phase, you capture verified email addresses. Not form submissions. Not retargeting pixels. Actual contact information from people actively researching your inventory.

You can identify 25-30% on the low end and up to 50% on the upper end of anonymous website visitors. Even at only 30%, that’s a massive jump from the 2-3% who fill out forms.

Dealerships that partner with early-adopter agencies generate more leads with the same budget. The ones stuck with less technologically advanced agencies? They’re burning money on traffic they can’t convert.

Bottom line: You’re missing 97-98% of potential leads by only capturing form submissions. Technology can identify 25-50% of anonymous visitors with verified email addresses, giving you 10-25x more contacts during the critical research phase when 70% of buying decisions happen.

3. Are you remarketing to those specific people, or are you just running retargeting pixels to a general audience?

There’s a huge difference between following people around with generic ads and actually reaching them with personalized messages.

When you capture a website visitor’s email and track the specific pages they view—noting how much time they spend on a new SUV versus a used sedan—you can deliver highly targeted marketing for the exact car that caught their interest.

Real personalization isn’t slapping a first name in an email subject line.

It’s sending an automated, personalized email moments after a visitor browses a particular vehicle. A detailed spec sheet. A video walk-through of the car. Warranty details. Dealership reviews. Customer testimonials.

Then you reconnect through dynamic YouTube video ads showcasing that specific model. OTT commercials that appear on their streaming services. Cross-channel retargeting that bypasses iOS anti-tracking limitations—which killed visibility for over 75% of users after Apple’s update.

This immediate, relevant communication keeps your dealership top-of-mind and moves the customer further down the path to purchase.

Bottom line: Real personalization means triggering automated, vehicle-specific content across email, YouTube, and OTT based on actual browsing behavior—not generic retargeting ads with a first name. This bypasses iOS tracking limitations and reaches prospects on channels they actually use.

4. Can you show me which of our ad dollars directly led to someone walking into the showroom or picking up the phone?

Attribution is hard. I get it.

It’s not always possible to draw a direct line from a specific ad campaign to a customer walking into a dealership. But there’s a big difference between attribution being imperfect and attribution being ignored.

Here’s how I think about it:

Tier 1: Direct attribution when you can get it

Store Visit conversions in Google Ads
Call tracking with source-level data (not just “Google” but which campaign, which keyword)
UTM discipline on every URL so you know exactly which campaign drove the visit
SMART Leads AI matches anonymous site visitors back to showroom traffic

Most agencies don’t have clean UTM structures or consistent call tracking. That’s not an attribution problem. That’s a setup problem they’re calling an attribution problem.

Tier 2: Directional attribution when you can’t

If a dealer was generating 80 leads a month at $45 cost per lead last year and they’re now at 110 leads at $38, something is working. You don’t need to prove which specific banner ad the buyer saw in February.

The honest conversation sounds like this: “We can show you what we can directly track. We can show you what’s directionally true. And we’ll tell you clearly which is which. What we won’t do is dress up a guess as a metric.”

The real problem with “assisted conversions”? Agencies use them as a catch-all to claim credit for everything while being accountable for nothing. Every channel assisted. Nobody caused anything.

The question to ask any agency: if you removed that channel tomorrow, what would change?

If they can’t answer that, the attribution model is decorative.

Bottom line: Attribution doesn’t have to be perfect, but it can’t be ignored. Direct attribution (Store Visits, call tracking, UTM discipline) tells you what caused results. Directional attribution (year-over-year cost per lead trends) shows what’s working. “Assisted conversions” without accountability is decorative reporting. Ask: “If we removed this channel tomorrow, what would change?”

5. When something isn’t working, how fast do you adjust, and what data are you using to make that call?

In reality, when someone says “wait for more data,” they usually mean one of three things: they’re uncomfortable making a decision, they’re protecting a recommendation they already made, or they genuinely don’t know what to look for.

Statistical significance sounds rigorous. Sometimes it’s just a stall.

Here’s my actual framework for when to pivot vs. when to wait:

Pivot fast when:

You’re burning budget with zero engagement signals (low CTR, no view-through, no site activity)
The creative is the obvious problem, and you can see it in the data within the first 7-10 days
The audience targeting is clearly misaligned

Wait when:

You’re in a learning phase that legitimately needs volume (Meta’s algorithm needs impressions, Google PMax needs conversion data)
The offer or seasonality hasn’t had a fair window
You’re seeing mixed signals, not negative signals

For most mid-market campaigns, if something isn’t showing any life after two weeks and a reasonable spend threshold, you already have your answer.

You’re not waiting for more data. You’re waiting for courage.

Agility isn’t about being reckless. It’s about having a pre-agreed decision framework with the client before the campaign launches, so when the data shows a problem, nobody’s arguing about whether to act.

Bottom line: “Wait for more data” usually means discomfort, protecting past recommendations, or genuine confusion. Pivot within 14-21 days when you see zero engagement signals. Wait when you’re in a legitimate learning phase with mixed (not negative) signals. The difference between vendors and partners is having a pre-agreed decision framework before launch.

What Should You Agree On With Your Marketing Agency Before Launch?

Before we launch anything, I sit down with the client, and we answer four questions together:

1. What does success look like at 30 days?

Not “we want more leads.” Actual numbers. Cost per lead target, click-through rate floor, conversion rate benchmark. If you can’t define it before launch, you’re just guessing afterward.

2. What’s our kill threshold?

If we hit day 21 and spend X dollars with zero meaningful engagement, we’re pulling the creative and rebuilding. We agree on that number in advance. No debate when we get there.

3. Who has final say on changes?

This one exposes a lot. Sometimes the client says, “Just run it by me.” Then you find out their approval chain is three people and a legal review. That’s not agility. You need to know the actual decision-making structure before a problem shows up.

4. What are we NOT willing to change?

Every client has a sacred cow. A logo placement, a tagline, a channel they love for emotional reasons. Get it on the table early so it doesn’t become the thing they protect when results are bad.

Most agency-client relationships skip this because agencies are afraid the client will say, “this sounds like a lot of work before we’ve even started.”

But the clients who push back on that conversation are usually the ones who become the hardest to work with later. The ones who lean in? Those become real partnerships.

The honest version of “we just launch and hope for the best” is: the agency is optimizing for a signed contract, not a successful campaign.

That pre-launch conversation is what separates vendors from strategic partners.

Bottom line: Pre-launch conversations that define success metrics, kill thresholds, decision authority, and non-negotiables separate strategic partnerships from vendor relationships. Agencies that skip this are optimizing for signed contracts, not campaign success.

Are Third-Party Automotive Vendors Wasting Your Budget?

This accountability gap isn’t just about digital agencies. It’s the entire vendor stack.

I typically see this around third-party vendors like CarGurus, Autotrader, Cars.com. GMs throw their budget on all three (if not more) and cannot truly estimate which one is delivering leads and customers.

There’s overlap. There’s cannibalization. The ad spend is not efficient.

The mentality is: if I spend my budget on everything third-party, I’ll cover all my bases. But they don’t realize that less is more in this case. A lot of GMs do not calculate their Cost Per Lead with each vendor. They only look at the overall ad cost and how many cars are sold.

To accurately test performance, pause one vendor at a time for 30 days.

If your lead volume drops during this period, resume that vendor the following month and proceed to test a different one. This process will help you determine which vendors are providing the most valuable leads for your business.

Monitor conversion rates and cost per lead during this testing period to assess the effectiveness of each vendor. Track key metrics like lead volume, conversion rates, cost per lead, and overall ROI.

Most GMs don’t know what would happen because nobody ever structured their marketing spend to be testable. Everything runs simultaneously, budgets never change, and the agency reports on all of it as if it’s one big success.

You can’t isolate variables that were never isolated.

Bottom line: Multiple third-party vendors (CarGurus, Autotrader, Cars.com) create overlap and cannibalization without clear ROI tracking. Test by pausing one vendor at a time for 30 days and monitor lead volume, conversion rates, and cost per lead. Less is often more when you can’t isolate what’s working.

How to Evaluate Your Marketing Agency

If your current agency can answer all five of those questions with real numbers and specifics, they’re doing a solid job. Stick with them.

If they can’t, or if the answers are vague, that’s worth thinking about.

Because those aren’t trick questions. That’s just what data-driven marketing looks like in 2026.

At New Road, those questions are our entire foundation. SMART Leads AI answers questions one and two. Our omnichannel stack handles three. Attribution tracking covers four. And five is just how we operate—every decision we make comes from the data, not assumptions.

I’m not trying to trash whoever you’re working with. But I’ve talked to a lot of dealers who are spending $15-25K a month and can’t get straight answers on where that money is actually going.

If that sounds familiar, it’s probably worth a conversation.

“I don’t know” is the most honest thing a GM can say. The problem is that most agencies have trained them to never ask the question.

When a GM says that, I’m not hearing weakness. I’m hearing someone who’s been sold dashboards instead of clarity for years.

Good. That means we have an opportunity. Let’s find out together.

Why Dealerships Should Regularly Evaluate Their Marketing Agency

Marketing should never operate on autopilot. The digital landscape changes quickly, consumer behavior shifts, and advertising platforms constantly adjust their algorithms and tracking capabilities.

What worked six months ago may already be outdated. That is why dealerships should regularly evaluate their marketing agency’s performance using clear metrics like cost per lead, attribution accuracy, campaign agility, and lead quality. A strong agency welcomes this level of accountability because it forces both sides to focus on measurable results rather than vanity metrics.

When dealerships periodically review their marketing strategy and vendor performance, they protect their advertising budget, uncover inefficiencies, and ensure their marketing partner is actively driving showroom traffic instead of simply maintaining campaigns.

In a market where every advertising dollar matters, consistent evaluation is not skepticism. It is smart management.

Frequently Asked Questions

What percentage of website visitors can actually be identified without filling out a form?

Technology exists to identify 25-50% of anonymous website visitors with verified email addresses. This is a massive improvement over the 2-3% who traditionally fill out forms. The identification happens during the silent research phase when 70% of buying decisions are made, allowing dealerships to engage prospects before they’ve completed their research.

Why can’t most marketing agencies identify anonymous website visitors?

Most agencies lack the agility to adopt new visitor identification technology quickly. This isn’t a limitation of available tools—the technology exists. Agencies that haven’t implemented these solutions are either unaware of the capability, haven’t invested in the infrastructure, or are comfortable with traditional metrics that don’t require this level of accountability.

How is personalized remarketing different from standard retargeting?

Standard retargeting follows users with generic ads across the web. Personalized remarketing tracks specific pages viewed (like which vehicle models and how long), then triggers coordinated messaging across email, YouTube, and OTT streaming platforms with content specific to that vehicle. This includes spec sheets, video walkthroughs, warranty details, and testimonials—all automated based on browsing behavior. It also bypasses iOS anti-tracking limitations that killed visibility for over 75% of users.

What’s a reasonable timeframe to pivot a failing marketing campaign?

For most mid-market campaigns, you should see engagement signals within 14-21 days. If you’re seeing zero engagement (low CTR, no view-through, no site activity) after two weeks and a reasonable spend threshold, you have your answer. The exception is when you’re in a legitimate learning phase that needs volume (Meta impressions, Google PMax conversion data) or seeing mixed signals rather than purely negative signals. The key is having a pre-agreed decision framework before launch that defines the kill threshold.

How do I test which third-party vendor (CarGurus, Autotrader, Cars.com) is actually delivering results?

Pause one vendor at a time for 30 days while keeping everything else constant. If the lead volume drops, resume that vendor the next month and test a different one. Track lead volume, conversion rates, cost per lead, and ROI during each test period. Most GMs spread budgets across all vendors, thinking it provides coverage, but this creates overlap and cannibalization without clear attribution. Testing reveals which vendors justify their costs.

What should I expect from my agency regarding attribution—direct proof or directional trends?

Both, with clear labeling of which is which. Direct attribution includes Store Visit conversions in Google Ads, source-level call tracking (not just “Google” but which campaign and keyword), UTM discipline on every URL, and tools like SMART Leads that match site visitors to showroom traffic. When direct attribution isn’t possible, directional attribution (year-over-year cost per lead trends) shows what’s working. What you shouldn’t accept is “assisted conversions” as a catch-all that claims credit for everything while being accountable for nothing. Ask: “If we removed this channel tomorrow, what would change?”

What are the four questions I should answer with my agency before launching any campaign?

1) What does success look like at 30 days? (Specific numbers: cost per lead target, CTR floor, conversion rate benchmark)

2) What’s our kill threshold? (At what day and spend level do we pull a failing campaign?)

3) Who has final say on changes? (What’s the actual approval chain?)

4) What are we NOT willing to change? (Identify sacred cows early so they don’t protect bad results later.)

Agencies that skip this conversation are optimizing for signed contracts, not successful campaigns.

Does “wait for more data” always mean the agency doesn’t know what to do?

Not always, but often. It can mean three things: discomfort making a decision, protecting a recommendation they already made, or genuinely not knowing what to look for. Statistical significance sounds rigorous, but is sometimes just a stall. Legitimate reasons to wait include learning phases that need volume, offers that haven’t had a fair seasonal window, or mixed signals (not purely negative). The agencies that say “wait” most often are usually the ones that never defined success clearly at the start.

The reality is simple. If your agency cannot answer these questions to ask your marketing agency, they are not managing your marketing strategy. They are managing reports.

Dealerships spending $20K to $50K a month on marketing deserve clarity, attribution, and measurable growth.

Key Takeaways

Up to 98% of website visitors remain anonymous, but technology exists to identify 25-50% with verified email addresses—a 10-25x improvement over the 2-3% who fill out forms.
By the time a shopper fills out a form, 70% of their research is complete. Identifying anonymous visitors during the silent research phase lets you influence decisions before they’re made.
Real personalization triggers automated, vehicle-specific content across email, YouTube, and OTT based on browsing behavior—not generic retargeting with a name in the subject line.
Attribution should include both direct proof (Store Visits, call tracking, UTM discipline) and directional trends (year-over-year cost per lead). “Assisted conversions” without accountability is decorative reporting.
Pivot failing campaigns within 14-21 days when you see zero engagement. “Wait for more data” is often code for discomfort, protecting past recommendations, or confusion.
Pre-launch conversations that define success metrics, kill thresholds, decision authority, and non-negotiables separate strategic partnerships from vendor relationships.
Multiple third-party vendors (CarGurus, Autotrader, Cars.com) create overlap and cannibalization. Test by pausing one at a time for 30 days to identify which actually delivers ROI.