Make the Mailer Measurable: GoHighLevel for Traditional Agencies Going Digital

You have something most digital agencies will never build. Twenty, thirty, sometimes forty years of relationships in a regional market, client families you have worked with across two generations of owners, and a reputation that means the phone still rings when a business in your town needs marketing. That is not a legacy asset in the polite sense of the word. It is a moat.

What you do not have is an answer to the question that has started arriving in every review meeting. The client's new marketing coordinator, or the founder's daughter who took over operations, or the private-equity operator who bought your client last year, asks what the campaign generated. Not what it cost. Not how many households it reached. What it generated.

And there is no answer, because there has never needed to be one.

This piece is about closing that specific gap, and it is deliberately not about becoming a digital agency. The pivots that destroy traditional agencies are the ones that start by telling a thirty-year direct-mail shop that its craft is obsolete. The pivots that work start somewhere much less dramatic — with a phone number.

Why does the measurement problem feel like an existential threat when it is really a plumbing problem?

Because the question sounds like a verdict when it is actually a request for data. When a client asks what the mailer produced and you cannot say, the silence gets interpreted — by them and, worse, by you — as evidence the mailer produced nothing. It almost certainly did not produce nothing. It produced results that nobody instrumented.

Here is the actual mechanics of the situation. A direct-mail drop of 12,000 pieces goes out on a Tuesday. Over the following six weeks the client's phone rings more than usual. Some of those callers mention the mailer, most do not. A few walk in holding it. The receptionist writes nothing down, because writing things down was never part of the job. At the quarterly review the client says business was decent in the spring but they are not sure the mail is worth it any more, and the agency has no counter-argument except thirty years of judgement.

That is not a marketing failure. It is a data-capture failure, and it sits entirely in the last three feet of the process — the moment a human being decides to respond and there is nothing in place to record which stimulus caused it.

The reason this feels existential is that your younger competitors do not have better media. They have dashboards. A three-year-old digital shop pitching your client can show a screen with numbers on it, and your forty years of market knowledge loses to a screen, not because the screen is smarter but because it answers the question you are not answering.

Three numbers worth holding in mind. Somewhere between sixty and eighty percent of the response to a well-targeted local direct-mail programme arrives by phone rather than by web form, which means an agency measuring only web forms is measuring the minority of its own results. Roughly a third of mail-driven calls arrive more than thirty days after the drop, which means an agency using a digital-style seven-day attribution window is discarding results it earned. And a call is worth several times a form fill in most local service categories, because a person who dials has already decided to buy.

You are not losing on performance. You are losing on the scoreboard.

What actually changes when a traditional campaign becomes measurable?

Everything downstream of the campaign and nothing upstream of it. The creative does not change. The media plan does not change. The list does not change. What changes is that the campaign now has a unique response path attached to it, and that path writes to a database.

The comparison below is the single most useful thing to put in front of a client, because it makes the change concrete without implying the previous decade was wasted.

Look at the last two rows, because they are where the retainer lives. Once the calls are recorded and the contacts are captured, you discover things that were previously invisible and are frequently worth more than the campaign itself. Missed calls are the classic example. A regional client running a mail programme will routinely miss ten to fifteen percent of inbound calls — lunch hours, after five, the Saturday after a drop — and every one of those is a person who chose them and got voicemail. An automatic text back within sixty seconds recovers a meaningful share of them, and it is the easiest win you will ever hand a client.

That is the whole strategic move in one sentence. Attribution does not just prove the old media works. It surfaces the leaks that only become visible once you are counting, and fixing leaks is monthly work.

How do you set up call tracking for print, radio and direct mail without breaking anything?

Start from the principle that the client's main line is sacred and never changes. Everything you add is additive.

Call tracking means assigning a distinct phone number to a distinct marketing source, publishing that number on that source only, and forwarding it to the destination that always answered. The tracking system sits invisibly in the middle, logging the time, the caller's number, the duration, the recording where permitted, and the source, then passing the call through with no perceptible delay.

The build sequence that works, in order.

Decide the number map before provisioning anything. One number per decision, not one per printed piece. If you would never drop the newspaper insert independently of the mailer, they can share a number. If you might drop one of two radio stations, they need separate numbers. A typical regional client stabilises at five to seven numbers. Write the map down as a table with four columns — channel, number, campaign name, in-market dates — because in eight months somebody will ask what a number was for and memory will not hold it.

Use local numbers, not toll-free, for regional traditional media. Response rates in regional markets are measurably better against a local area code, and toll-free reads as a call centre in exactly the towns where your clients trade on being local. Reserve toll-free for clients with genuine multi-state reach.

Match the number's format to the medium. For radio, get a number that can be read aloud twice in a spot without stumbling — repeated digits and clean pairs. A number with a memorable pattern will materially outperform a random one in a thirty-second read, and this is one of the few places where you should be willing to pay for a specific number rather than accepting whatever the pool offers. For print, legibility matters more than memorability because the reader has the number in front of them. For outdoor, the number needs to be readable at speed, which usually argues for a QR code or a vanity URL instead.

Set the forwarding and the fallback in the same sitting. Each tracking number forwards to the client's main line. Configure what happens if that line does not answer within a set number of rings — a second destination, a mobile, then voicemail with an automatic follow-up text. Test the whole chain from an outside phone before it goes anywhere near artwork. Test it again the morning of the in-home date.

Handle recording law explicitly and in writing. Call recording consent rules vary by state, province and country, and several US states require all-party consent. Either enable a recording announcement, restrict recording to jurisdictions where single-party consent applies, or turn recording off and keep only call metadata. Metadata alone — time, duration, source, caller number, answered or missed — is enough for attribution. Recording is a coaching tool, not an attribution requirement. Put the decision in the client contract rather than in an email.

Route the data into the CRM immediately, not eventually. A tracked call that lands nowhere is a statistic. A tracked call that creates a contact record with the campaign name on it is an asset you can nurture, report on and re-market to for years. This is the hinge between "we bought a call tracking tool" and "we built a measurement system", and it is the reason to run this inside a platform like GoHighLevel rather than bolting a standalone tracking product onto a client's existing phone setup. When the call, the contact, the pipeline stage, the follow-up text and the report all live in the same place, an account manager with two hours of training can run it. When they live in four places, they cannot.

Never recycle a number that appeared on a durable asset. Mailers get kept. Fridge magnets last a decade. A number printed on anything with shelf life should be forwarded permanently, even after the campaign ends. Numbers are cheap; a lost call from a three-year-old mailer is not.

What is dynamic number insertion, and when does a traditional agency actually need it?

Dynamic number insertion, almost always written as DNI, is a small script on a web page that changes the phone number shown to a visitor based on how they arrived. Someone landing from the QR code on the mailer sees one number, someone arriving from a search result sees another, and both calls get attributed correctly even though there is only one page.

For a traditional agency the honest answer is that you do not need it on day one, and telling a client otherwise is how you build a system your own staff cannot maintain.

The simple architecture — a genuinely separate landing page per campaign, each with a hard-coded unique number typeset into the page — is more robust, easier to explain, impossible to break with a caching plugin, and perfectly adequate up to about three concurrent campaigns. Most eleven-person agencies should run this for the first year.

DNI earns its place when three things become true at once. You are running more than three or four campaigns into overlapping web destinations. You want the client's main website, not just campaign landing pages, to attribute traffic by source. And someone on your team is comfortable enough with a tag manager to install and verify a script. Until all three are true, the added complexity buys you nothing except a new failure mode.

When you do implement it, two practical guardrails. First, verify the swap on a real mobile device from a real QR scan rather than trusting a desktop preview, because that is the path ninety percent of your traffic will take and it is the path most likely to be cached. Second, keep a static fallback number in the page markup so that if the script fails to load the visitor still sees a working number. A page that renders no phone number at all because a script timed out is a far worse outcome than a call attributed to the wrong campaign.

How should you structure unique URLs and QR codes on printed pieces?

Treat the URL as part of the creative, not as a technical afterthought, because on print it is read by a human who has to type it or point a camera at it.

Vanity URLs need to survive being read aloud and typed by hand. The working standard is a short, pronounceable path on the client's own domain, all lowercase, no hyphens if avoidable, no numbers that could be a digit or a word. Something like clientdomain.com/spring beats clientdomain.com/dm-2026-q2-a every time, even though the second is more informative internally. Keep the informative version as the internal campaign name and map it to the human-readable path with a redirect. Register the obvious misspellings as redirects too — people will type the plural, the singular and the one with the missing letter.

Use the client's domain, not a shortener. A third-party shortener on a printed piece looks like spam, cannot be fixed if the service disappears, and throws away the trust the client's own brand carries in that market. This matters more in regional markets than anywhere else.

One URL per campaign, one per variant if you are testing. If a mail programme has two creative versions going to two list segments, they need two paths. If it has one creative going to one list, it needs one. Resist per-piece uniqueness — personalised URLs at the individual recipient level are a real technique but they roughly triple the production complexity and they are not where a first-year measurement programme should spend its budget.

QR codes carry the same URL with a tracking parameter appended. The code should resolve to the vanity path plus a source parameter identifying the medium, so a scan and a typed visit both land on the same page but report separately. This tells you something genuinely useful — whether your audience in that market scans or types — which changes how you design the next piece.

Design the code to actually get scanned. Minimum printed size around two centimetres square for a handheld piece and considerably larger for anything read at distance. High contrast, dark code on light ground, never reversed out. Adequate quiet zone around the edge. And a four-word instruction beside it telling the reader what they get, because a naked QR code on a mailer has a fraction of the scan rate of one captioned with something like "scan for your quote".

Print-test the code before the run. Scan it from a printed proof, not from a PDF on screen, using two different phones. Codes fail in production for mundane reasons — a code placed across a fold, a code trapped in a varnish, a code printed too small in a resized layout — and every one of those failures is discovered by the client rather than by you if you skip this step.

Give the landing page one job. The page a mail recipient lands on should carry the same offer and the same visual language as the piece that sent them, a phone number, a short form, and almost nothing else. A campaign landing page that dumps the visitor onto the client's general homepage discards most of the response and all of the attribution.

What attribution window should traditional media use, and why is the digital default wrong?

An attribution window is the period after a campaign runs during which a response still counts as caused by that campaign. Choosing it is a judgement call, and for traditional media the digital defaults are actively misleading.

Direct mail is a physical object that persists in the home. A recipient reads it, puts it on the counter, and acts when the need arrives — which might be nine weeks later. Radio works by frequency and memory, so response accrues across a flight and continues after it ends. Outdoor is a repeated impression whose effect is almost entirely delayed. None of these behave like a paid search click.

Practical windows that hold up in front of a sceptical client.

Three rules make windows defensible rather than arbitrary.

Set the window before the campaign runs and write it into the reporting template. A window chosen after the results are in is not a measurement, it is an argument.

Measure from the in-home date, not the mail date. A drop that hits the post on the fourth and lands in homes on the ninth has a day-one that is the ninth. Getting this wrong shifts your entire response curve and makes the campaign look slower than it was.

Never shorten a window retroactively to flatter a number, and never extend one either. The credibility you are building is the product. An agency that reports honestly on a mediocre quarter is trusted on the good one.

One more thing worth doing from the first campaign. Record the response curve by day, not just the total. After three or four campaigns for the same client you will have something genuinely valuable and completely proprietary — a known response shape for that client in that market, which lets you tell them on day six whether the drop is tracking ahead or behind. That is a level of insight no dashboard-first competitor can match, because it requires knowing the client's business, and knowing the client's business is what you have always been good at.

How does Brennan and Sons prove this works?

Brennan and Sons is a thirty-four-year-old regional print and direct-mail agency with eleven staff. Second-generation ownership. A client list heavy on home services, regional healthcare, funeral homes, agricultural equipment dealers and a credit union, most of them ten years deep or more. Excellent creative. Genuinely good list work. Zero measurement.

The pressure arrived the way it usually does. Their largest client, a home services company running four mail drops a year at around forty thousand dollars of annual spend, brought in a new operations director who asked for campaign performance data. Brennan and Sons sent reach and frequency figures. The operations director replied asking what it generated. There was a month of silence, then a request for proposals that included two digital agencies.

What was built, and in what order.

Month one was a single campaign, not a platform. One client, one upcoming mail drop, seven tracking numbers, three landing pages, one pipeline. The deliberate decision was to instrument the campaign that was already scheduled rather than propose a new programme, because the client had not agreed to anything yet and the agency needed evidence more than it needed a contract.

The number map for that first drop looked like this. One number on the mailer itself. One on the companion newspaper insert. One each on two radio stations running a supporting flight. One on the client's van livery, treated as permanent from day one. One on the existing yard signs. One held as a control for the client's own organic and word-of-mouth calls, which turned out to be the most informative number of the seven.

The landing pages were three — one for the mail QR code and vanity URL, one for the radio's spoken URL, one for the insert. Each carried the campaign offer, a form, and the matching tracking number hard-coded into the page. No DNI in year one, by choice.

The attribution window was set at sixty days from in-home date, agreed with the client in writing before the drop, because the client's average job value sat above four thousand dollars and the purchase was clearly considered.

What the first drop produced.

Twelve thousand two hundred pieces, in-home the second week of the month. Over the sixty-day window the instrumented campaign recorded 212 tracked calls, 47 form submissions and 31 inbound texts, against roughly 14,000 dollars of combined media and production.

The breakdown mattered more than the total. 141 of the 212 calls came from the mailer number. 38 came from the two radio numbers combined, split heavily toward the morning-drive station. 21 came from the insert. 12 came from the van. The control number, which nobody expected to be interesting, logged calls from people who had heard about the offer from a neighbour — a measurable word-of-mouth halo that the client had never seen quantified and which became a permanent part of every subsequent report.

Response by day showed the shape the agency had always suspected but never proven. Forty-one percent of calls arrived in the first eleven days, a long shallow middle through week five, then a distinct second bump at days thirty-two to thirty-eight that lined up exactly with the radio flight's second week. Under a fourteen-day window, that second bump would not have existed.

The client's own close data, matched back against the tracked contacts, produced the number that ended the conversation. Forty-one closed jobs traceable to the campaign, and 180,000 dollars of revenue inside the ninety-day reconciliation. Against 14,000 dollars of spend.

What the measurement surfaced that nobody had been looking for.

Twenty-eight of the 212 calls went unanswered. Nineteen of those hit between five and seven in the evening, four on the Saturday after the drop, five during a weekday lunch. Every one was a person holding the mailer who had decided to call.

Brennan and Sons added a single automation — an automatic text within sixty seconds of any missed call, offering a callback slot. Over the following two drops that automation recovered a little over a third of missed calls into booked conversations. It took an afternoon to build. The client noticed it faster than they noticed the attribution reporting, because it produced money they could feel.

The commercial outcome, which is the actual point.

The RFP was cancelled. The home services client moved to a monthly retainer of 1,200 dollars covering number management, campaign configuration for each drop and flight, the missed-call and nurture automations, and a monthly report.

Over the following twelve months Brennan and Sons converted nine project-based clients onto monthly retainers, at an average of roughly 740 dollars per month. That is close to 80,000 dollars of new annual recurring revenue against setup work billed at about a thousand dollars per client — and, more importantly, against a client base they already had. They did not win nine new clients. They stopped being paid only when something was printed.

The client roster did not change. The relationship did. And the mail volumes went up, not down, because for the first time the client could see what the mail was doing.

Why does staff adoption fail, and what actually fixes it?

This is where these projects die, and it is worth being blunt rather than reassuring.

Your longest-serving people are not resisting because they are stubborn or old or afraid of computers. They are resisting because the new system does something specific and unpleasant to them — it converts twenty years of accumulated, unwritten, high-status expertise into a set of fields on a screen that a twenty-six-year-old can fill in faster than they can. The account director who has known the client's founder for fifteen years watches a junior produce a report in four minutes that they cannot produce at all. That is not fear of technology. That is a correct assessment of a status threat.

Treat it as a structural problem and it becomes solvable. Treat it as an attitude problem and you will lose people you cannot replace.

Give it a named internal owner with real authority. Not the owner's son because he is under forty. The person on staff who is genuinely curious, regardless of their title — often a mid-level account manager or the traffic coordinator, because they already live at the intersection of everything. Give them ownership of the configuration, time carved out of their billable week, and the explicit right to say no to feature requests. A system with no internal owner reverts to the old process within a quarter, every time, no matter how good the build is.

Roll out by role, not by department. Nobody needs to learn the whole platform. Segment the training by what each person actually touches. The account manager needs to read a report and update a pipeline stage. The media buyer needs to request numbers and check response curves. The receptionist needs to know that some calls are tracked and that nothing about answering them changes. The studio needs to know that a number and a URL must be confirmed before artwork goes to proof. That is four different training sessions of twenty to forty minutes each, not one two-hour session that everybody sits through and nobody absorbs.

Deliberately exclude people who do not need it. This is the counterintuitive one and it does more for adoption than any amount of encouragement. If your senior designer never needs to open the CRM, do not give them a login. Forcing universal adoption creates a population of people who have accounts they do not use, which makes the system look unused, which makes everyone else conclude it did not take. Fewer, more committed users is a stronger foundation.

Write one page per role, not a manual. A single side of paper with the four or five things that person does, with screenshots, in the agency's own language using the agency's own client names. Fifty-page documentation is written to make the person who wrote it feel thorough. Nobody reads it. A laminated one-pager beside the monitor gets used daily.

Make the first visible output flatter a sceptic. Choose the first report to be about the account of your most senior, most resistant account director, and let them present it to their own client. The moment the system makes them look prescient in front of a relationship they have owned for a decade, the internal argument is over. This is the highest-leverage single decision in the whole rollout and it costs nothing.

Say out loud that the old skills are the scarce ones. Anybody can be taught to read an attribution dashboard in a week. Nobody can be taught in a week how to negotiate a regional radio buy, or which list source actually performs in your county, or how to keep a family-owned client for eighteen years. Name that difference explicitly and repeatedly, because the unspoken fear in the room is that the new system makes the old skills worthless, and the truth is precisely the opposite — the measurement layer is a commodity, and the judgement it feeds is not.

Expect a six-to-ten-week dip and plan for it. Productivity genuinely falls during a rollout. If you have not budgeted for that, the dip gets interpreted as proof the system does not work, usually in week three, usually by the person who was against it. Say in advance that weeks two through eight will feel slower and that this is expected. Naming it in advance converts a crisis into a milestone.

What does the ongoing retainer actually cover, and what is it worth?

The setup makes the case. The retainer is the business model, and it has to be built out of work that genuinely recurs rather than work you invented to justify an invoice.

The honest inventory of monthly work.

Number and campaign administration. Provisioning numbers for new drops and flights, retiring or permanently forwarding old ones, keeping the number map current, maintaining forwarding rules and after-hours routing. Small each month, non-optional, and the thing that quietly breaks if nobody owns it.

Campaign configuration per cycle. Every drop, flight or insert needs its landing page, its QR code, its source parameters, its in-home date entered and its window set. For a client running four drops and two flights a year this is six configuration events, each a few hours.

Automation maintenance. The missed-call text back, the form-response sequence, the nurture track for people who called but did not buy, the review request after a completed job. These need reviewing and adjusting as the client's offers change.

Monthly reporting and the conversation attached to it. The report itself should be largely automated. The half-hour conversation about what it means is the part the client is actually paying for, and it is the part your competitors cannot deliver because they do not know the client's market.

Reconciliation of tracked responses to closed revenue. Quarterly rather than monthly for most clients, and the highest-value thing on this list. Matching tracked contacts back to the client's own job or sales records is what converts "212 calls" into "180,000 dollars", and it requires a relationship with the client's operations people that you already have and a digital-first competitor does not.

Pricing that holds up in regional markets. Setup near one thousand dollars, covering the number map, landing pages, forms, pipeline, automations, reporting build and role-based training. Monthly retainer between 400 and 1,200 dollars, with the bottom of that range for a single-channel client running two campaigns a year and the top for a multi-channel client with high call volume and quarterly reconciliation.

The arithmetic that matters to your agency. Nine clients at an average of 740 dollars a month is 80,000 dollars of annual recurring revenue that arrives whether or not anything goes to press this quarter. For an eleven-person agency whose revenue has always been project-shaped and seasonally violent, that is not a new revenue line. It is a different kind of business.

And it is defensible in a way project work never was. A client who has twelve months of attribution history living in a system you built, tied to campaigns you configured, reconciled against revenue you helped them measure, does not switch agencies over a proposal. The switching cost you were never able to create with creative quality alone, you can create with data continuity.

How do you sequence the first ninety days without betting the agency on it?

Do not launch a digital division. Instrument one campaign for one client and let the results decide the rest.

Weeks one and two — choose and map. Pick a single client with an upcoming traditional campaign, a decent relationship and, ideally, mild scepticism about their own results. Build the number map with them. Agree the attribution window in writing. Decide the landing page structure. Nothing has been provisioned yet and nothing has been promised.

Weeks three and four — build and test. Provision numbers, build landing pages, generate and print-test QR codes, configure forms and pipeline stages, set the missed-call automation, wire notifications to the right people. Test every number from an outside phone. Scan every code from a printed proof on two devices. Get the studio to confirm the number and URL on the artwork proof before it goes to press, using the sign-off sheet they already use.

Week five — train by role. Four short sessions. Account manager, media buyer, front desk, studio. One page each. The internal owner runs these, not you and not the software vendor, because the person who trains becomes the person people ask afterwards.

Weeks six through fourteen — run and watch. The campaign goes live. Record response by day from the in-home date. Resist the urge to report early; a partial curve looks like a failure. Fix the small things that will inevitably break — a forwarding rule that misses an after-hours case, a form notification going to someone on holiday.

Week fifteen — report, then sell. Present the full window with the response curve, the channel split, the missed-call recovery and, if the client's records allow, the revenue reconciliation. Then make the retainer offer in the same meeting, while the number is on the screen. This is the only moment where the offer is obvious rather than pushy.

Weeks sixteen onward — templatise and repeat. Save the entire configuration as a reusable template — pipeline stages, automations, form structures, landing page layouts, report format, the one-page training documents. The second client should take a third of the time the first one took. By the fourth or fifth, one person can deploy a full attribution build in a week, and the economics of a one-thousand-dollar setup start working properly.

The clients to approach in what order, once you have the first proof. Start with the ones already asking the measurement question, because they have pre-sold themselves. Then the ones whose campaigns are largest, because the absolute numbers will be most impressive. Then the ones with the longest tenure, because they will say yes out of trust and you must be certain the system works before you spend that trust.

What should you refuse to do?

A short list, because the failure modes here are as important as the method.

Do not rebuild the client's website as part of this. It is a separate project with a separate budget and a separate set of arguments, and folding it in will delay the attribution build by three months and give the client a reason to say no to both.

Do not promise digital media performance you cannot deliver. Adding attribution to traditional media is a different competence from running paid social. If the client asks you to run ads because you now seem technical, either build that capability deliberately or partner, but do not improvise it inside an engagement whose credibility you are still establishing.

Do not let the reporting become a vanity dashboard. A report with forty metrics is a report nobody reads. Five numbers — responses by channel, cost per response, missed-call recovery, pipeline value, revenue attributed — carried consistently month after month, will do more for renewal than any amount of visual sophistication.

Do not instrument a campaign whose creative you would not defend. Attribution is neutral. It will prove a weak offer is weak with the same clarity it proves a strong one is strong. Fix the offer first if it needs fixing, because the first tracked campaign sets the client's expectation of what tracking means.

Do not describe any of this to the client as digital transformation. They do not want to be transformed. They want to know what the mailer did. Answer that question, keep answering it every month, and the transformation happens without anybody having to be told they were doing it wrong for thirty years.

Where does this leave a traditional agency in five years?

In a considerably stronger position than the panic suggests, provided the pivot is framed correctly.

The competitive assumption of the last decade — that measurable channels would replace unmeasurable ones — turned out to be half right. Measurable channels won. But the reason traditional media kept working in regional markets is that it never stopped working; it just stopped being counted, and being uncounted got mistaken for being ineffective.

The agencies that survive this are not the ones that abandoned print for paid social. They are the ones that kept the craft, kept the relationships, kept the market knowledge that takes decades to accumulate, and added the one missing layer — a number, a URL, a code, and a database behind them.

That layer is not hard. It is a few weeks of setup and a monthly discipline. What is hard is the thing you already have, which is thirty years of a regional market's trust, and a client base that would rather buy measurement from you than buy relationships from someone else.

Start with one campaign. Prove the mailer works. Everything else follows from that.