Payment Links That Actually Get Paid in a Medical Practice

Payment links are the easiest part of patient collections to implement and the part practices most often blame when collections do not improve. The link works. Patients can use it. What determines whether a balance gets paid is almost entirely upstream of the link: when it arrives, whether the patient understands the number attached to it, and whether anyone was available to answer the question they had about it.

Most practices adopt payment links as a channel swap, replacing a paper statement with a text message, and then measure the result as a channel experiment. It usually produces a modest lift and then plateaus, because the statement was never the binding constraint. The constraint is that a meaningful share of patients who receive a balance do not know what it is for, do not believe it is right, or cannot pay it in full and were never offered another option. A link does not address any of those.

Patient balances are frequently the residue of a claims process the patient never saw, and that process produces more friction than practices tend to assume. KFF found that across healthcare.gov insurers with complete data, nearly 19% of in-network claims were denied in 2023, with issuer denial rates ranging from 1% to 54%. Of those denials, about 16% were for an excluded service and 9% for lack of preauthorization or referral.

What matters here is not the denial rate itself but what it produces downstream. A denied or partially paid claim frequently becomes a patient balance, and the patient receives it without context, often months after the visit. From their side it is an unexplained number for something they believed was covered.

KFF also found consumers appealed about 1% of denied in-network claims in 2023. Patients overwhelmingly do not appeal. What they do instead is not pay, or call the practice, and the practice absorbs a conversation created by a payer decision it did not make and cannot see from the statement.

So the first thing a collections workflow has to do is not send a link. It is to make the balance legible.

Timing beats channel, and the window is short

The strongest predictor of whether a patient balance gets paid is how long it has existed.

A balance communicated within days of adjudication, while the visit is still a real memory, gets paid at a materially different rate than the same balance at ninety days. By then the patient has lost the context, the practice has spent three statement cycles, and the conversation has changed from paying a bill to disputing an old one.

This is why the channel swap disappoints. Practices replace a paper statement at day thirty with a text at day thirty and measure a small improvement, having left the actual variable untouched. Moving first contact from day thirty to day three usually does more than every channel and copy change combined.

The second timing rule is about sequencing against other contact. A balance message that arrives before a post-visit survey turns the survey into a referendum on the bill. One that arrives during an active clinical episode, between a procedure and its follow-up, reads badly and generates calls. Both are avoidable with a simple rule about what else is in flight for that patient.

The plan has to be offered before the patient asks

Most practices have payment plans and treat them as a concession, available on request, usually after a balance has aged and someone has called to complain.

That design guarantees the plan reaches the patients who complain rather than the patients who need it, and those are different groups. The patients who most reliably do not pay are the ones who look at a balance they cannot cover, decide nothing good happens by engaging, and stop opening the messages.

Offering a plan at first contact, automatically, above a threshold the practice sets, changes the decision that patient is making. It moves it from pay this or ignore it to choose between these two, and the second framing produces engagement even when the patient picks the full payment.

Setting the threshold is a practice decision and worth making deliberately rather than defaulting. It should reflect the balance size at which your own collection rate starts falling off, which is visible in your own data and is usually lower than administrators expect. Below it, offer the link. Above it, offer the link and the plan in the same message, with the monthly amount already calculated so the patient is not being asked to do arithmetic before deciding anything.

Stored payment methods belong in the same conversation, with consent handled explicitly. A card kept on file with clear terms, stated at the time and confirmed, is a materially different thing from one captured by default, and the difference is worth being careful about.

Someone has to answer the question

The reason balance outreach stalls is rarely that patients refuse to pay. It is that they have one question and no cheap way to ask it.

What was this charge for. Why did insurance not cover it. I thought I paid this. This is for my daughter’s visit and I have not seen anything about mine. Every one of those is answerable from records the practice already holds, and every one of them, unanswered, converts into a balance that ages.

A text message with a link is a one-way channel, which is why so many practices see high delivery and low conversion. The patients who would have paid after a two-minute explanation instead do nothing, and the practice reads that as unwillingness.

This is squarely administrative work and it automates well. An outbound call that states the balance and what it is for, answers the common questions from the account, offers the link or the plan, takes the payment or sends it, and hands to a person the moment the caller disputes the charge or asks something the account does not answer. Disputes go to billing staff, always, because a disputed balance is a claims question rather than a collections one.

Measure the funnel, not the collection rate

Collection rate is a lagging composite that hides where the failure is, and practices steering by it usually cannot tell an outreach problem from a payer problem.

Break it into stages. Share of balances with a first contact within a defined window of adjudication, which is the timing variable and the one most likely to be quietly broken. Contact-to-open rate. Open-to-payment rate. Plan uptake rate above the threshold. Share of contacts that resulted in a dispute routed to billing, which is a quality signal about the balances themselves rather than about collections.

That last one deserves attention. A rising dispute rate is usually not a collections failure at all. It points at a coding, eligibility or verification problem upstream, and it is one of the few places where the collections workflow can see a defect that originated at the front desk weeks earlier.

Then one comparison worth running deliberately: payment rate by days from adjudication to first contact, on your own data. It is a cheap analysis, it will show a curve, and the curve ends the internal debate about whether timing matters more than the message.

Key Takeaways

The link is not the constraint. Patients who do not pay usually do not understand the balance, do not believe it, or cannot pay it in full and were never offered an alternative.
Balances often arrive already disputed because a payer decision the patient never saw produced them. Make the balance legible before asking for payment.
Move first contact from day thirty to day three. That single change usually beats every channel and copy adjustment combined.
Sequence balance messages against other patient contact. A bill arriving before a post-visit survey turns the survey into a referendum on the bill.
Offer the plan at first contact above a practice-set threshold, with the monthly amount already calculated. Plans available only on request reach the patients who complain, not the ones who need them.
Measure the funnel: contact-within-window, contact-to-open, open-to-payment, plan uptake, and dispute rate. A rising dispute rate is an upstream coding or verification defect, not a collections failure.

Payment links are worth having and they are not a collections strategy. What moves patient collections is contacting people while the visit is still recent, telling them plainly what the balance is for, offering a plan before they have to ask for one, and making sure the single question standing between them and paying can be answered by somebody within a couple of minutes. The link is simply where that conversation ends. Practices that install it and change nothing else generally find out that the statement was never the problem.

Sources

A large share of balances arrive already disputed

Timing beats channel, and the window is short

The plan has to be offered before the patient asks

Someone has to answer the question

Measure the funnel, not the collection rate

Key Takeaways

Related reading

Ready to See It in Action?