Should Your Pharmacy Offer Delivery? What the Call Data Says

Ask most pharmacy owners whether their customers want delivery and you will get a shrug. Some tried it, saw little uptake, and quietly stopped. Others assume the demand is there but cannot size it.

We found an unusual way to measure it. When we put an AI voice agent on an Australian pharmacy’s phone line and started capturing what callers actually asked for, delivery came out as the single largest category of request. Not scripts. Not opening hours. Delivery.

This guide covers what that data showed and why most pharmacies underestimate delivery demand. It also covers how to size it for your own catchment, and how to run delivery without wrecking your dispensary workflow.

The short answer

Delivery demand at a community pharmacy is mostly arriving by phone, and at a pharmacy with a poor answer rate, most of it is never recorded.

In our data, delivery requests were 40.2 percent of all captured phone requests, more than any other category. At the same pharmacy, before measurement, 74.9 percent of inbound calls went unanswered.

Those two numbers together are the whole argument. If your delivery trial “failed”, check whether anyone was answering the phone during it.

What the call data showed

Across a limited beta at one independent Australian community pharmacy in 2025, an AI voice agent answered calls and converted them into structured tickets for staff. Of the calls it handled, 57 percent produced an actionable request.

Here is the breakdown of those requests.

Two caveats before anyone over reads this. First, it is 112 requests at one pharmacy over a limited window, so it is a signal about where phone demand concentrates, not a market study. We would not extrapolate it to a national delivery market.

Second, and more important, these requests come only from the calls the agent actually handled. The agent ran during limited active periods rather than across all traffic, so this is not necessarily a representative slice of every caller. The request mix across the full call population could differ. The full methodology and limitations are published with the underlying report.

What it does establish, at least at this site, is that delivery is not a fringe request. It was the most common thing people rang up to ask for.

Why pharmacies underestimate delivery demand

Here is the trap, and it is a genuinely easy one to fall into.

Most pharmacies assess demand for a service by looking at how often it is requested. That works when you hear every request. It fails badly when you do not.

At the pharmacy we audited, roughly 99.6 calls a day went unanswered. Some of those will be the same person redialling, but each one is still a request that never got recorded. If delivery requests are distributed across all callers rather than only the persistent ones, then for every delivery request the pharmacy heard, roughly three went unheard.

A pharmacy in that position can run a six week delivery trial, count the requests it received, and conclude that interest is thin. The trial did not measure demand. It measured demand that survived a 25 percent answer rate.

This is the difference between a service failing and a service never being visible. Before you conclude your customers do not want something, check whether they could reach you to ask for it. Our guide to running a call audit on your own phone line walks through the method.

Why delivery demand is growing anyway

Independent of our data, the structural case for pharmacy delivery in Australia is getting stronger.

Australians make more than 443 million visits to community pharmacies each year across more than 6,000 locations. That is an enormous base of people for whom the trip is routine, and a meaningful minority for whom it is difficult.

Getting to care is already hard for a lot of people. In 2024-25 the ABS recorded 26.6 percent of Australians delaying or not seeing a GP when they needed to. For patients with limited mobility, the pharmacy trip sits in the same category of friction. The AIHW reports that around 1 in 3 emergency department presentations were classified as lower urgency, with rates 3.6 times higher in remote areas than in major cities.

Meanwhile the economics of dispensing alone have tightened. Grattan Institute analysis puts pharmacy remuneration at about $15.90 per script in 2024-25 in today’s dollars. The Guild reports prescription volumes remain flat after the move to 60 day dispensing.

Flat script volume, tight per script margin, and a customer base that increasingly struggles to get anywhere. Delivery sits directly in that gap. We cover the wider set of options in our checklist of alternative revenue streams for independent pharmacies. Delivery is also one of four places pharmacies quietly lose customers: how to get more customers into your pharmacy.

Sizing the opportunity for your own pharmacy

Do this before you buy a single insulated bag.

1. Pull 30 days of call records. Ask your phone provider for inbound calls with timestamp, duration and disposition. Note your answer rate.
2. Sample the answered calls. Listen to or review a sample and tag what each caller wanted. You are looking for the share asking about delivery, getting something dropped off, or asking whether you can bring it out to the car.
3. Adjust for what you missed. If your answer rate is 60 percent, you heard roughly 60 percent of the requests. Do not treat what you heard as the total.
4. Map the catchment. Count how many of your regulars live within a 10 minute drive. Drive time, not kilometres.
5. Identify the priority cohorts. Aged care residents, people with mobility limitations, parents with young children, and anyone on a regular repeat cycle. These are the customers for whom delivery changes behaviour rather than just adding convenience, and the ones whose medication adherence is most at risk.
6. Estimate the round, not the drop. Delivery economics are set by how many drops fit in one run, not by the cost of a single drop. Ten deliveries in one loop is a different business from ten deliveries on demand.

If that exercise produces fewer than a handful of viable drops per run, the answer is probably no, or not yet.

How to run pharmacy delivery without breaking the dispensary

The failure mode for pharmacy delivery is not lack of demand. It is that it quietly consumes staff time that was already fully committed.

Batch the runs

Fixed delivery windows beat on demand every time. One or two runs a day, at set times, lets you build a route and lets customers plan around it. On demand delivery turns every request into an interruption, which is the exact problem you are trying to solve.

Capture the request properly at the point of contact

A delivery request that arrives as a sticky note is a delivery request that will go wrong. You need the address, the contact number, the preferred window, the items, and any handover requirement, captured in one structured place.

This is where the phone problem and the delivery problem meet. If requests arrive by phone during the 11am and 1pm peaks, and nobody can reach the handset, the request never enters your system at all.

Set the free radius by drive time

Distance is a bad proxy for cost. Five kilometres across a city grid can take longer than fifteen on an open road. Price the radius on the time it consumes.

Decide the handover rule before you start

Who can receive the medication, what happens when nobody is home, whether anything requires signature or identity check, and what your cold chain rule is. Write it down and train to it, rather than deciding on the doorstep.

Check your jurisdiction’s requirements

Delivery of scheduled medicines in Australia is governed by state and territory poisons legislation, and the requirements are not uniform across jurisdictions. Confirm your own state or territory health department requirements and current Pharmacy Guild guidance before you begin, particularly around Schedule 8 medicines, record keeping and handover.

This guide is about operations and demand. It is not regulatory advice, and you should not treat it as such.

Measure completion, not just capture

Taking the request is the easy half. In our beta, staff completed 84.8 percent of the requests captured. That is the number worth tracking, because a captured request nobody fulfils is worse than one you never took. It creates an expectation and then breaks it.

Delivery models compared

Model	Best for	Watch out for
Staff run, batched	Dense catchments, existing spare capacity in quiet hours	Staff time is real cost even when it feels free
Staff run, on demand	Very low volume, high value or urgent cases	Turns every request into an interruption
Third party courier	Wider catchments, unpredictable volume	Per drop cost, chain of custody, and less control of the patient experience
Hybrid: batched in house plus courier overflow	Most independents with growing demand	Two sets of rules to train staff on

Most independents that make delivery work start with one batched staff run a day inside a tight radius, and only widen once the round is consistently full.

Where the phone fits

The connective point in all of this is that delivery is a phone mediated service. People ring to ask for it, ring to check where it is, and ring to change the window.

That puts a pharmacy with a poor answer rate in an awkward spot. Launching delivery increases inbound call volume, into a phone line that is already the bottleneck. Done in that order, delivery makes the phone problem worse.

The sequence that works is the other way around. Fix the capture first, then launch the service. If you can reliably answer and structure inbound requests, delivery becomes an operations problem, which is solvable. If you cannot, it becomes a customer service problem, which is not.

On the options for fixing capture, see our guide to AI voice agents and missed calls at Australian pharmacies, including the section on when they are the wrong answer.

Where Krepko fits

Krepko builds Emily, an AI voice agent for Australian pharmacies. She answers the phone, captures requests as structured tickets your team can action, and passes clinical or complex calls to a pharmacist.

The delivery finding in this article came out of that work, and we were not looking for it. We put an agent on a phone line to see how many calls it could answer. The largest thing callers turned out to be asking for was whether the pharmacy could bring their medication to them.

That is the honest reason we think this is worth writing about. Delivery demand at that pharmacy was real, it was the biggest single category of request, and before anyone was answering the phone properly, essentially none of it was visible.

If you want to size your own numbers first, start with the Australian Pharmacy Phone Report and the audit method in it. If you want to talk to us after that, we are at krepko.com.au.