If you bake, arrange, cook or pack orders and then get them to customers yourself, email marketing for small business is the cheapest way you have to get a second order out of someone you already delivered to once. You have their address. You know what they bought. You know when they are likely to want it again. No other channel hands you that.
Most owners in this position never use it. They open an account, send a newsletter in March, and let it sit. The work that actually pays is duller and mostly happens once: collect addresses at the moments customers are already talking to you, switch on four or five automated messages, and settle on a send rhythm you can keep during a busy week.
This post is the whole playbook in order: list, automations, calendar, numbers, tools. Where a piece of it deserves its own post, you’ll find a link: specific local campaigns you can send this month, the wording for each individual email, and which platform to actually pay for.
The Bottom Line
- Automated flows produced about 41% of email revenue from 5.3% of sends in Klaviyo’s 2026 benchmark set — roughly $1.94 per recipient against $0.11 for one-off campaigns (Geysera, citing Klaviyo, retrieved 2026-09-24).
- Set up in this order: list capture, welcome series, order and delivery confirmations, win-back, then regular campaigns. The first four run without you.
- Average open rate sits near 19-20% across industries and a click rate of 2-5% is normal (ClickMinded, retrieved 2026-09-24). Your own list should beat that, because your subscribers have met you.
- Delivery details are marketing. Confirmations, delivery windows and “your order is on the way” messages get opened far more than any promotion you will ever write.
Metrobi drivers are rated 4.97/5
Trusted by local businesses for:
- Background-checked professionals
- Specialized in business deliveries
- Same drivers for consistency
- 4.97/5 average delivery rating
Why email pays better than social for a delivery-based business
Email’s advantage is structural rather than creative. A post on social reaches whoever the algorithm decides to show it to. An email reaches everyone who gave you an address, and you can aim it at the twenty people on one delivery route rather than at everyone.
The return figures quoted around email marketing, commonly $36 to $40 per dollar spent, come from large aggregate studies and are best treated as directional rather than as a forecast for your shop (ClickMinded, retrieved 2026-09-24). The specific economics that matter to you are narrower and easier to check: a customer who already ordered once costs you nothing to reach again, and a second order on a route you are already driving carries almost no extra delivery cost. That is where the margin is.
Three things make email unusually well suited to businesses that deliver:
- You know the delivery day. A Tuesday route means a Sunday email. Timing you could never get right on social is trivial here.
- Reorder cycles are short and regular. Bread, flowers, prepared meals and produce all run on repeat purchase. Email is the only cheap channel built for reminding.
- Operational messages carry marketing weight. The email confirming a delivery window is the one your customer definitely opens. What else sits in it is up to you. If you would rather those windows arrived as texts, an email-to-text service can push them to a phone without you running two systems.
How to build an email list when customers meet you at the door
List building is the step people skip, and nothing downstream works without it. You do not need a lead magnet or a pop-up strategy. You need to stop letting addresses go uncollected at the four or five points where customers already give you information.
- At checkout, online and off. An emailed receipt is the least awkward way to ask. Online, a single checkbox at checkout is enough; in person, offer to email the receipt instead of printing it.
- On the order form. Any business taking orders by form already collects an address to confirm the order. One sentence of consent turns that into a subscriber.
- On the delivery itself. A card in the box with a short URL or QR code converts better than almost anything, because the customer is holding your product when they read it.
- From wholesale and standing accounts. The café buying your pastries weekly has a manager who wants to know about supply changes. That is a list segment, not just a contact.
- From the people who ask about delivery. Anyone who asks whether you deliver to their street has told you they are interested and given you a location. Capture both.
Two rules keep this out of trouble. Ask permission plainly, because CAN-SPAM and GDPR both hinge on consent and on a working unsubscribe link. And record the neighbourhood or postcode with the address. That field is what later lets you email only the routes you actually serve.
The five automated emails to switch on before you write a single newsletter
This is where the money is, and it is the part that keeps working while you are driving. In Klaviyo’s benchmark data, abandoned cart is the single highest-yielding automation at about $3.65 revenue per recipient, with a welcome series at $2.65, browse abandonment at $1.07 and win-back at $0.84, against $0.11 for a standard campaign (Geysera, citing Klaviyo, retrieved 2026-09-24). Those figures come from ecommerce brands, so treat them as a ranking of what to build first rather than as revenue you should expect.
Build them in this order:
- Welcome series (2-3 emails). Who you are, what you make, which days you deliver where. Send the first within minutes of signup, while interest is high.
- Order confirmation and delivery window. Operationally necessary and the most-opened message you own. Include what arrives, when, and what to do if nobody is home.
- Abandoned cart or abandoned order form. One reminder a few hours later, a second the next day. If your order form is a Google Form rather than a cart, the same logic applies to anyone who starts and does not finish.
- Post-delivery follow-up. Sent a day or two after arrival. Ask how it was, make reordering one click, and this is where reviews come from.
- Win-back. Triggered by silence — 60 or 90 days without an order, whatever a normal gap looks like for your product.
Each of these needs actual wording, and the wording matters more than the trigger. The fill-in-the-blank copy for all five is laid out separately so you can paste rather than draft.
How often to email customers who order every week
The honest answer is that frequency depends on your reorder cycle, not on a rule. A bakery whose customers buy weekly can email weekly without complaint. A florist selling mostly for occasions cannot, and should not try.
A practical starting cadence for most delivery businesses:
- Automations: always on, triggered by behaviour. These are not part of your frequency budget — a customer who just ordered expects to hear about that order.
- One regular campaign a week or fortnight: what is available, what changed, what is in season. Tie it to your delivery calendar so it lands the day before cut-off.
- Seasonal pushes, three or four a year: the holidays that actually shift your product.
Watch unsubscribe rate rather than your own nerves. Mailchimp’s industry figures put a normal unsubscribe rate around 0.15% (ClickMinded, retrieved 2026-09-24). If yours stays near that, you are not sending too much, whatever your instinct says. If it climbs, cut frequency before you blame the content.
Segmenting by route solves the other half of the problem. If you deliver to three areas on three days, three smaller emails will beat one general one, and nobody gets told about a cut-off that does not apply to them.
Local campaign ideas worth putting on the calendar
Once the automations run, the recurring question is what to put in this week’s send. Short version: things happening near your customers and things happening in your kitchen or workshop. The photographs you take on the round are the raw material for most of it, and the good ones beat stock imagery every time. Owners often end up printing them as canvas prints for the shop wall as well as using them in email. New items, a market you’ll be at, a route you just added, a seasonal batch, a neighbourhood you’re expanding into, a partnership with the café down the road. If that partner is a coffee shop, their own coffee shop marketing strategies will tell you what they need out of the swap.
There’s a longer bank of these, with the trigger and timing for each, in the local email marketing ideas post, including which ones to segment by area and which work as a single send.
Email marketing metrics that tell you it is working
Four numbers are enough. Check them monthly, not daily.
| Metric | Typical benchmark | What it tells you |
|---|---|---|
| Open rate | ~19-20% average; above 25% is strong | Whether your subject lines and sender name are recognised |
| Click rate | 2-5% depending on industry | Whether the email gave people something to do |
| Unsubscribe rate | ~0.15% | Whether your frequency and relevance are sustainable |
| Bounce rate | Under 2-3%; above 5% is a problem | Whether your list is clean and your sending reputation is intact |
Benchmarks from ClickMinded and Campaign Monitor ranges cited there, retrieved 2026-09-24.
Two cautions on open rate. Apple’s Mail Privacy Protection inflates it by pre-loading images, so the absolute number is soft. Use it to compare your own sends against each other, not against the industry. And revenue per email sent is the number that actually decides anything. If your platform reports it, that is the metric to watch; if it does not, count orders that arrive within 48 hours of a send.
Choosing email marketing tools for a small delivery operation
Almost any platform will do the job at your list size, and switching later is annoying but survivable, so this decision deserves less agonising than it usually gets. What matters for a delivery business specifically: automation on the free or cheap tier, a form that can capture a postcode or neighbourhood, segmentation you can actually use, and an integration with wherever your orders come from.
Rough shape of the market at a small list: free tiers exist but most are tight. Mailchimp’s and Omnisend’s free plans cover 250 subscribers, MailerLite’s 250, EmailOctopus and Sender 2,500, and Brevo prices on email volume rather than contacts with 9,000 monthly sends free (EmailToolTester, retrieved 2026-09-24). At around 1,000 contacts, paid entry plans cluster between roughly $13 and $19 a month.
The full comparison, covering what each one costs as the list grows, which have automation on the cheap tier, and which to pick for a wholesale-plus-retail mix, is in the email marketing services breakdown. Pick from there, then come back and build the five automations above.
Your first 30 days, in order
- Week 1: Pick a platform. Turn on address capture at checkout, on your order form, and with a card in the box.
- Week 2: Write and switch on the welcome series and the order/delivery confirmations.
- Week 3: Add the post-delivery follow-up and the abandoned-order reminder. Import any addresses you already hold, with consent you can point to.
- Week 4: Send your first regular campaign. Segment it by delivery day. Note the four metrics above so you have a baseline.
Then the win-back flow, once you have enough history to know what “gone quiet” means for your product.
Frequently asked questions
How many subscribers do I need before email is worth doing?
A few hundred is plenty, and a list of 50 regular buyers is already worth emailing. Percentages matter less than order value here. Twenty subscribers who each order $40 of bread weekly are worth more than a thousand idle ones.
Is a free plan enough to start?
For a few hundred contacts, usually yes, with one caveat: check that automation is included rather than reserved for paid tiers, because the automated flows are the part that pays.
Do delivery confirmations count as marketing emails?
They are transactional, which is why they reach people who ignore promotions and why consent rules treat them differently. Keep the promotional content in them light, because a confirmation stuffed with offers loses the trust that makes it get opened.
What should I do with an old list I have not emailed in two years?
Do not blast it. Send a short re-permission email to a small batch first, remove anyone who bounces, and only scale up if the bounce and complaint rates stay low. Importing a stale list in one go is the fastest way to damage your sending reputation.
