Why your traffic is fine and your inquiries are not

Every owner I talk to knows their traffic number. Almost none of them know their conversion rate. That gap is not laziness. Analytics dashboards put sessions on the front page in a big font, and inquiries are buried three clicks deep in a report nobody opened.

So the instinct when leads go quiet is to buy more traffic. More ads, more posts, more keywords. That is the most expensive available answer to a problem that usually costs nothing to fix.

Is this a traffic problem or a conversion problem?

Look at whether both numbers moved together. If sessions dropped and inquiries dropped in step, you have a traffic problem. If sessions held flat and inquiries fell, you have a conversion problem, and no amount of new visitors will cover it.

The second case is more common than owners expect, because search visibility is stubborn. Rankings tend to decay slowly. Websites break fast. A form plugin updates, a phone number stops linking on mobile, a redesign moves the contact button below three screens of photography, and the traffic curve does not move at all.

I ran restaurants for sixteen years before this, and the pattern is the same as a dining room that is full but not making money. Covers are fine. Check average is not. Counting the covers harder does not fix the check average.

How do I calculate my conversion rate?

Divide the number of real inquiries by the number of sessions in the same period, then multiply by 100.

Real inquiries means completed contact forms, phone calls that actually connected, booked appointments, and messages you can trace back to the site. It does not mean clicks on the phone number, newsletter signups, or anything a bot can trigger. If you count soft actions, your conversion rate will look excellent and your bank account will disagree.

Use one full month at minimum, and a quarter if you can. Small sites bounce around too much week to week for the number to mean anything, which is the same reason split testing rarely produces a usable answer at local traffic levels.

Two rules that keep the number honest. Use the same date range for both figures. And exclude your own visits, plus any traffic from a campaign that sent people who were never going to buy.

That is the whole calculation. The hard part is not the arithmetic. It is having a tracking setup that counts inquiries at all, which is why getting conversion tracking in place before you spend a dollar matters more than any dashboard you buy afterward.

What does the calculation look like on real numbers?

Can I trust the numbers themselves?

Less than you think, and this is the part almost nobody checks.

Two examples from the same 2014 KDD paper by Ron Kohavi, Alex Deng, Roger Longbotham and Ya Xu, written out of Microsoft and LinkedIn, are worth carrying around. In 2013 Bing moved its content delivery network, and over the following months several teams reported that clickthrough rate was rising and abandonment was falling. It turned out the new network improved click tracking fidelity, not user behavior. When they replaced beacon based tracking with redirects, click loss for some browsers dropped by more than 60%. A large part of the celebrated improvement was an artifact of counting.

The second one is funnier and more common. MSN shipped a new autosuggest algorithm and searches referred to Bing rose dramatically. Investigation showed the new code was issuing two searches whenever a user picked a suggestion, one of which the browser discarded.

Your version of this is smaller and just as real. A tag firing twice on the thank you page. A phone click counted as a call. A form event that stops firing after a theme update, which is the one I find most often on local sites, because it makes inquiries appear to collapse while the phone keeps ringing.

Before you diagnose a page, verify the counter. Submit your own form, place a real call, and confirm both appear where you expect.

Why does more traffic not fix this?

Because conversion rate is a multiplier, and multiplying a broken number gets you a bigger broken number.

Say your site turns two visitors in a hundred into inquiries. Doubling traffic doubles inquiries, which sounds fine until you price it. New traffic costs money every month, forever. Fixing the page costs money once. The site is the only asset in your marketing that keeps paying after you stop paying for it.

There is also a ceiling. A local business in a defined service area cannot buy unlimited relevant traffic. There are only so many people searching for a plumber in your city this month. Once you are visible for the terms that matter, the growth left in the system is on the page, not above it.

This is the same reason the ad question depends on your conversion rate rather than your ad budget. The math that decides whether paid traffic can work for you starts with the percentage you just calculated, not with the click price.

What is actually stopping people from inquiring?

Usually one of four things, and none of them are subtle once you look.

The visitor cannot tell what you do or where you do it. They landed from a search, they get a few seconds of attention, and your headline says something about passion and quality instead of the service and the city.

The visitor does not believe you yet. Reviews sit in the footer, license numbers are missing, and the photos are stock. The trust elements that actually change behavior are specific and verifiable, and most sites carry the decorative version instead.

The visitor cannot act without effort. The form asks eleven questions, the phone number is an image, or the booking link opens a login screen.

Or the visitor has a question you refused to answer. Price is the big one. People who cannot find a range assume they cannot afford you, and they leave without ever telling you.

Where does this argument break down?

Three places, and the third is the one that costs people a year.

When the inquiry arrived and nobody caught it. CallRail's analysis of 1.1 million de-identified conversations, published January 2025, put missed call rates at 32% in health care, 28% in legal, 14% in home services and 9% in real estate. CallRail sells call tracking and its sample is its own customers, which if anything makes those figures flattering. A missed call is not a conversion problem. It is a staffing problem wearing one.

When the loss is happening off your website entirely. Sterling Sky's 2026 field data found clicks to call from Google Business Profiles declining even for profiles holding stable rankings, across 179 profiles at 34 US law firms over two years, while website clicks did not fall the same way. The decline is mobile specific and it is the call button. If most of your inquiries never touch your website, your site conversion rate can be perfect while your phone goes quiet.

When your traffic is genuinely tiny. Below roughly a hundred sessions a month, the percentage swings on one extra call and tells you nothing. Count leads, watch for a cliff, and put your attention on visibility.

What to do this week

Pull last quarter's sessions and last quarter's real inquiries. Do the division. Write the number down, because you now have a baseline and every change you make from here is measurable against it.

Run step three from the table above with your own numbers, so you know which movements your traffic can actually detect and which ones you should ignore.

Then open your own website on your own phone, from a coffee shop connection, as if you had never seen it. Time how long it takes you to find the phone number and tap it. If that takes more than a few seconds, you found your first fix.

Third, check that you control the analytics account, the domain, and the site files. Owners who do not control their own domain and site data usually discover it at the exact moment they need to change something quickly.

Repeat the conversion calculation quarterly. One number, same method, same window.

Questions about your conversion rate?

Email me at eric@seod.com with two numbers: your sessions last quarter and your real inquiries last quarter. I will do the division, run the detection arithmetic to tell you whether any movement you are seeing is readable at your volume, and name the single thing I would look at first. No form, no call required.

I answer these myself. If the honest answer is that your site is converting fine and your problem is that not enough people are finding you, I will say that, and it will save you a redesign you did not need.

Or keep reading the rest of the conversion library.