How to run a re-engagement email campaign on a list nobody has emailed in years.

"We don't have a marketing budget." The owner of a 29-year-old insurance agency said it on my first day, while his CRM held almost 9,000 contacts nobody had emailed in years. I spent the first couple of weeks understanding that list, sorted it into five groups and ran the first re-engagement email campaign there, a plain email from the agent they already knew about the one policy they were missing. Existing clients had been calling about twice a week before I got there, and within a few weeks it was 17 calls a week.

Months later, the owner was the one who brought up budget, and the list is the reason. That is the list-first rule: prove it with what they have before you ask for what they don't. It works best before the next budget is set, while a result can still shape it.

Why the list you already own is the first place to look.

Philipp Schmitt, Bernd Skiera and Christophe Van den Bulte tracked about 10,000 customers of a German bank for almost three years (Journal of Marketing, 2011), and referred customers were worth at least 16% more than similar customers who weren't referred. Those were consumers, and B2B referrals may differ, which group 3 below lets you check with long-time clients.

What an old list can do to your sending reputation.

ZeroBounce, which sells email verification, found that 23% of the more than 11 billion addresses its customers checked in 2025 were invalid. Those are lists people chose to check, which makes the figure a warning more than a decay rate for your list.

Spamhaus describes a second risk. Some abandoned addresses hard bounce for a while, "often 12 months or more," and are then turned back on as spam traps.

"You should NEVER mail old addresses unless you can identify where the address came from and the specific purchase that address is associated with." Laura Atkins, Word to the Wise, 2017

Atkins added in 2022 that a verification service removes bad addresses and can't tell you which former customers will report your email as spam.

How to run a re-engagement email campaign in five steps.

1. Sort the list into groups by relationship.

Engagement filters don't help on a list nobody has emailed. HubSpot's "Don't send to unengaged contacts" setting counts the sends after the last open or click, and a contact who has received nothing has nothing to count. The setting is on by default, and it still drops anyone with no open or click on their last 11 emails (16 if they engaged before), however old those sends are, and a group can shrink at send time.

1. Current clients with one active product and no second one.
2. Clients with a renewal or contract end date in the next 90 days.
3. Likely referrers: current clients with the longest history with you.
4. Past clients with no active product and no email from you after they left.
5. Contacts with a quote or inquiry date and no purchase.

Build each group in HubSpot under CRM > "Segments" (formerly lists), in Salesforce as a report added to a campaign, in Salesforce Marketing Cloud Account Engagement as a static list, in Mailchimp as a tag, in Constant Contact as a list, or in a spreadsheet with "Group" and "Source" columns. Name each one with the year, like "2027-q1-renewals-90-days."

At a larger company, agree on who owns each group before anything goes out, and suppress contacts in an open deal or an active sequence.

2. Check where each address came from.

Add columns for source (purchase, inquiry, event, import), the date of the last purchase or inquiry, and opt-out status. Remove opt-outs, anything bought, rented or scraped, and any address you can't tie to a purchase or an inquiry. Mailchimp's terms also forbid purchased and third-party lists.

Platform rules can be stricter than the law. HubSpot requires a verifiable opt-in record for every contact in a marketing email, and in HubSpot, contacts without one stay out of the email and get a call from someone they know instead. Mailchimp counts purchase history as permission under its terms and recommends a reconfirmation email for contacts you haven't emailed in some time.

In the US, CAN-SPAM makes "no exception for business-to-business email." In Canada, CASL allows implied consent for two years after a purchase and six months after an inquiry. At a larger company, legal signs off on the groups, and EU and UK contacts get their own review.

3. Verify, then send in small batches, newest first.

Verify the list and remove invalid addresses, plus role addresses (admin@, postmaster@) unless one is a client's own contact on record. Set stop rules before the first batch. HubSpot can suspend marketing email at a 5% hard bounce rate, 0.1% spam reports or 3% unsubscribes in a month. Google asks senders to stay below 0.1% spam in Postmaster Tools and never reach 0.3%, Yahoo's line is 0.3%, and Google recommends starting with a low volume and increasing it slowly.

At 100 to 300 emails a day, Postmaster Tools may show nothing, because data "might be missing if the total number of messages for a given day is too low," and that's why you watch your platform's spam reports and unsubscribes, plus "who is this?" replies. Gmail doesn't send single complaints back to your platform (its feedback loop gives high-volume senders aggregate data in Postmaster Tools), and the spam reports you see are a floor.

Start with the newest records and work backwards, as Atkins recommends, in a separate send from your regular newsletter. Google requires SPF, DKIM, DMARC and one-click unsubscribe from anyone sending close to 5,000 messages a day to personal Gmail accounts, and Outlook.com rejects mail from domains over 5,000 a day unless it passes SPF, DKIM and DMARC.

At a larger company with a dedicated IP, the batches also avoid the "sudden volume spikes" Google warns against, watched daily in Postmaster Tools and Microsoft SNDS.

4. Write one plain email per group, from a person they know.

Navdeep Sahni, Christian Wheeler and Pradeep Chintagunta ran randomized field experiments with three companies and millions of recipients (Marketing Science, 2018). In the main experiment, the recipient's name in the subject line raised opens from 9.05% to 10.80% and sales leads from 0.39% to 0.51%, and cut unsubscribes from 1.2% to 1.0%. Those were consumer programs measured before Apple's privacy change, and B2B results may differ. The change cost nothing and moved leads, not only opens, and the authors saw the same effect when the goal was keeping customers who had bought before, which is why every subject line below carries the first name.

Atkins adds: reintroduce yourself, remind them of their history, make replying simple and limit the emails per address.

5. Count calls and replies, not opens.

For Apple Mail readers who turn it on, Mail Privacy Protection "prevents senders from seeing if you've opened the email message they sent you." Count replies, calls, meetings and closed deals per 100 delivered. In 1905, Sears customers in Iowa who handed out its catalogs "sent Sears the names of people who received the catalogs," the Sears Archives records, and earned a premium when those people ordered. Credit each group the same way, by the replies and calls it brings back.

Make sure replies reach someone that day, and the inbound lead follow-up guide covers that part.

A worked example with numbers.

Example numbers, for a made-up services firm and not the agency above: 9,000 contacts, untouched for three years. Removing opt-outs and addresses with no purchase or inquiry behind them leaves 7,400, and verification removes 1,300 more, leaving 6,100. Renewals (480) go first in batches of 120 a day, then current clients (2,100) in batches of 300, then lapsed clients, newest first.

Hard bounce rate = hard bounces ÷ emails sent, and spam rate = spam reports ÷ emails delivered (HubSpot divides by emails sent, Postmaster Tools by inbox deliveries). In a batch of 300, 15 hard bounces is 5%, and one spam report is 0.33%, which is why the first complaint means a pause. Responses per 100 delivered = (replies + calls) ÷ delivered × 100. If 470 renewal emails are delivered and bring 19 replies and calls, that's about 4 per 100.

Common mistakes and the pushback a specialist would raise.

"Send a reconfirmation email and drop non-clickers." HubSpot recommends that for subscribers who stopped opening, and Mailchimp for contacts you haven't emailed in some time. For clients your platform lets you email (in Mailchimp, purchase history counts), a reason to write (a renewal date, a missing product) usually earns more replies, and reconfirmation stays with the newsletter.

"Use a subdomain to protect the main domain." Google counts every message from the same primary domain toward the 5,000, subdomains included.

Making every contact a marketing contact. Set only the groups you'll email this period as marketing. Each counts toward your HubSpot tier at least until the next update date, and the groups need to fit the tier first.

A two-week test you can run on one group.

1. Pick one group, usually renewals in the next 90 days, take the 200 newest records and split them at random into two halves of 100 (a RAND() column in a spreadsheet works).
2. Email one half in two batches of 50, two days apart, with the stop rules above, and hold the other half back for the 14 days.
3. For 14 days, track delivered, hard bounces, spam reports, unsubscribes, replies, calls and meetings for both halves.
4. It's a result if the emailed half brings at least twice the replies and calls of the held-back half, and at least five from the emailed half, with hard bounces under half the platform line and no more than one spam report.
5. A result earns the next group, and a miss means changing the reason to write.

Halves of 100 catch a large lift, not a small one. The result gives you a number for the budget conversation. The monthly marketing report guide shows how to present it, and the prospect list guide applies the same idea to cold email.