Questions to ask ESP vendors before you sign

Seven questions, split by what they are for: three things you make a vendor prove while you can still watch, and four you get in writing while you still have leverage.

By Roy Dungo, Operations · Alium Research From hundreds of verified buyer interviews Interviews through August 2026 Updated August 25, 2026

Based on verified interviews with the lifecycle and email leaders who run these platforms, at DTC and enterprise brands. Buyers are anonymized before publication; vendor names and views are reported as given. No vendor paid to appear or could edit this page.

An ESP (email service provider) evaluation is less process-heavy than you might expect. Roughly half the ones in our interviews never run a formal RFP, and the ones that do converge on the same narrow gate: one or two finalists in a sandbox. What separates the buyers who report few surprises is not how much process they ran — it is that they made vendors demonstrate a short list of things rather than assert them, and put a second short list in the agreement rather than an email. For the hindsight version of this, the companion read is what ESP buyers wish they'd known.

What questions should you ask an ESP vendor before you sign?

Seven — three the vendor has to prove in a sandbox or trial, and four it has to commit to in writing.

Make them prove

  1. Can our own team build and QA a campaign we have already run, in your product?
  2. Can the platform model our real customer records and our brand or region hierarchy?
  3. What does the product show natively about inbox placement, reputation and rendering?

Get in writing

  1. What is the IP warm-up schedule at our list size and send volume, and the worst case?
  2. Which service tier does this price include, and who can open a ticket?
  3. What is your twelve-month record of unplanned downtime?
  4. What do one-year and three-year terms cost, and what flexibility do we give up?

Work backwards from your contract renewal date, which is what starts most of these evaluations in the first place.

Do you need an RFP to buy an email platform?

Often not. Roughly half the email-platform evaluations in these interviews never run a formal RFP, and formality tracks company size and deal size more than the category. One mid-market buyer describes the shape plainly: the process is informal at the start, especially for a tool like an email platform, and becomes formal only once it reaches evaluation and contracting. At a small brand, decisions land in days. At a multi-brand group, the parent may recommend a platform and the operating brand has to build a business case for an exception. At one furniture group, every new vendor requires a vote by a management committee — described by the buyer as an obstacle to buying anything at all.

Two shapes an email-platform evaluation takes. Roughly half the buyers in these interviews are on the right-hand one.
Stage The formal path The informal path
What starts it A renewal or a requirement enters a planning cycle; landscape review and requirements gathering, sometimes through a third party The renewal date arrives, or a bill does
Building a shortlist A shortlist demoed, with business and technical audiences separately at larger firms A short list assembled from peers, agencies and the buyer's own experience
The real gate One or two finalists into a sandbox. Narrow, and this is where the decision is actually made Demos, and sometimes a trial
Diligence Procurement, IT, legal and security, running in parallel Procurement and legal enter only at contracting
How the decision is made Pricing benchmark, scoring and term negotiation The judgment of the people who will run it

Which path you are on changes what these questions are worth, but not which ones matter. On the informal path nobody else will ask them for you — there is no procurement function to catch a missing support tier or an auto-renewal clause. The questions below are the same either way; the difference is who has to remember to ask.

What should you make an ESP vendor prove before signing?

Three things: campaign setup and QA, your real customer and organizational data structure, and native inbox-placement and reputation reporting. All three are answerable only by demonstration, in a sandbox or a trial, with your people driving — and the bar in every case is your own current numbers, not the vendor's benchmark.

Build something you have already built

Ask Give our team sandbox access and let us set up the campaign or test we ran last quarter.
Require Working access for your people rather than a guided walkthrough, and time-to-setup measured against how long the same work took you internally.
Red flag A sandbox offered only as a vendor-driven session, or only after signature.
Verify Have the same person QA and ship a second version a week later, unassisted. One buyer scored vendors on ease of QA specifically — it is the step that breaks first once you are in production.

Your record structure, not their reference customer's

Ask Here are our messiest records and our brand, region or business-unit hierarchy. Model them.
Require A specific answer on how subscriber records resolve when one person is two kinds of customer, and how permissions and content cascade across brands or regions.
Red flag Hierarchy described as achievable "with configuration," with no estimate attached.
Verify Change one journey and confirm the change propagates everywhere it should. A publisher whose customers buy both personally and on behalf of an organization — often from the same address — found the platform kept separate keys for each, producing duplication that surfaced only after migration.

What the product shows you about placement

Ask Show me, in the product, what you report on inbox placement, reputation and rendering — not opens and clicks.
Require A live view, plus an honest statement of where a specialist tool is still needed.
Red flag Deliverability described only as an outcome the platform is good at, with no monitoring surface to show you.
Verify Price the third-party monitoring you will still buy and add it to the quote before you compare quotes. Buyers run these tools alongside every platform — the question is how much of the picture you are paying for twice. Why deliverability stays yours either way is lesson two on the retrospective page.
This is the average buyer's picture. Which parts apply to your stack?A 15-minute interview returns the read specific to you — and what peers with your setup chose.
Get my personalized read →

What should an ESP vendor commit to in writing?

Four things: the IP warm-up schedule, the support model included in the price, twelve months of incident history, and one-year versus three-year pricing and terms. No demo settles these. They are cheap to ask now, expensive to raise once a preferred vendor exists, and each belongs in the agreement rather than in an email.

How long will IP warm-up actually take?

There is no single duration these interviews support — but buyers who have run one consistently report it taking longer than the vendor predicted.

Ask Give me the ramp schedule by week at our list size and send volume — and the actual durations of your last five migrations at our scale.
Require Five real durations rather than an average, a named worst case, what caused it, and what our sending looks like in weeks one through six.
Red flag A generic warm-up window, with no worst case and no evidence from migrations at your scale.
Verify Model the revenue gap across the ramp against your own peak-season calendar, then ask who is accountable if the schedule slips. Buyers who plan this well validate and clean the list immediately before warm-up begins, not after.

What support is included in this price?

Ask Is support self-serve, pooled, or named — how many of our people can open a ticket, and what does the next tier up cost?
Require Named roles, the ticket-eligibility rule, the escalation path and response targets, in the agreement.
Red flag "Full support" with no tier named. Buyers report discovering after signature that user support access was gated behind an upgrade.
Verify Ask a customer of your size — not the vendor — who they call and how long it takes. This is the question buyers on both sides of a decision keep landing on: one small brand rates its platform down purely on the service model rather than the features, and one organization is staying on an aging platform because the support is the best the buyer has seen.

The incident record

Ask Twelve months of unplanned downtime and maintenance windows, and how you notified customers each time.
Require The actual record, plus what a maintenance window costs a sender on a schedule like yours.
Red flag Uptime quoted as a percentage with no incident history behind it. One buyer shortlists on past outages explicitly, alongside scalability and client base.
Verify Confirm the notification practice with a current customer. In the complaints buyers describe, the damage usually came from how the outage was communicated rather than from the outage itself — one reports a maintenance window consuming a whole business day.

The term, and what the term buys you

Ask Price this at one year and at three, and tell me what the multi-year discount costs me in flexibility.
Require Both quotes, the auto-renewal and notice terms, and which capabilities sit behind a tier or add-on — named at signature rather than discovered at renewal.
Red flag A multi-year term presented as the only commercial option.
Verify Benchmark the price independently — one buyer hired an outside firm to do exactly that, and kept which vendor it preferred to itself throughout the negotiation. Then book your own review at six months rather than waiting for the renewal.

When should you start looking at a new email platform?

Start early enough to overlap the incoming and outgoing platforms while warm-up runs. In practice that means working backwards from the incumbent contract's renewal date rather than starting when a capability gap becomes urgent — which is also what these interviews describe: buyers time evaluations to a contract expiry, and one runs a phased cutover deliberately, winding the outgoing vendor down as its term expires while testing the replacement in parallel.

That overlap is the point, and it is the thing you lose by starting late. Warm-up runs for weeks, so keeping the outgoing platform live preserves margin during the cutover. A buyer who signs the month the old contract ends has removed their own. Some sequence deliberately for other reasons too — one retailer deferred its email evaluation until its customer-data project was finished, on the grounds that migrating onto a new platform with unresolved data would move the problem rather than solve it.

Want this read against your own stack?

Get my read →

The honest caveat

Buyers almost never credit their evaluation for a good outcome — the unhappy ones describe the platform, not the diligence. So treat this as a pattern rather than a promise. What the interviews support is narrower: the buyers reporting the fewest surprises are disproportionately those who made vendors demonstrate rather than assert, and who settled the service tier and the term in the agreement rather than in conversation. Two further limits worth stating. Nobody in these interviews describes a written scorecard for an email platform, so there is no weighting model here to lift. And the sandbox is the only in-evaluation test buyers describe in any detail — the seed-list tests and placement trials a page like this could easily invent are not in the corpus, and are not here.

Common questions

Do you need an RFP to buy an email platform?

Often not. Roughly half the email-platform evaluations in our interviews never run a formal RFP, and formality tracks company size and deal size more than the category. Where a formal process does exist, the decisive step is narrow: one or two finalists in a sandbox. The sandbox is the real gate, not the paperwork.

What should you test in an ESP trial or sandbox?

Three things: whether your own team can build and QA a campaign you have already run, whether the platform models your real customer records and brand or region hierarchy, and what it reports natively on inbox placement and reputation rather than opens and clicks. Use your team and your data, and measure against your current platform rather than the vendor's benchmark.

How long does an IP warm-up take when you switch email platforms?

There is no single duration these interviews support. Buyers consistently report warm-up taking longer than vendors predicted, so do not evaluate a vendor on a generic average. Ask instead for the ramp schedule by week at your list size and send volume, the actual durations of the vendor's last five migrations at your scale, and a named worst case — then model the revenue gap across the ramp against your own peak-season calendar.

What should you get in writing before signing an ESP contract?

Four things: the warm-up schedule with dates and a worst case; the service model, meaning whether support is self-serve, pooled or named and who can open a ticket; twelve months of incident history; and the price at one year as well as three, with the auto-renewal and notice terms named.

When should you start looking at a new email platform?

Early enough to overlap the incoming and outgoing platforms while warm-up runs — which in practice means working backwards from your incumbent contract's renewal date. Buyers who start late lose that overlap, and it is what lets you run both platforms in parallel until the new one is ramped.

Related reading

This page is the evaluation itself. For what buyers learned after the decision, what ESP buyers wish they'd known; for what puts a platform in play in the first place, why companies switch ESPs; and for the commercial picture behind the term question above, what buyers actually pay for an ESP at scale. The complaints that recur once you are inside a platform are graded on the top complaints about enterprise ESPs. The same format for another category is questions to ask CDP vendors, where the evaluation is more formal and the tests are run against your own data.

Get this research made for your stack

Which of these matter most depends on your list size, your sending calendar, and what you're migrating off. Do a 15-minute interview about your stack and get the version that applies to you — what buyers at your scale made vendors prove, what they put in the contract, and what they found out afterwards.

Get my personalized read — 15-min interview

No password · your interview is anonymized before it ever informs a page like this one.

Methodology. Alium conducts verified interviews with software buyers — the lifecycle and email leaders who run these platforms. This page draws on the email-platform interviews in that corpus, conducted through August 2026, focusing on how buyers ran their evaluations rather than on the platforms themselves. Where a recommendation is Alium's inference rather than something buyers reported, the page says so. Buyer identities are verified at interview time and anonymized before publication; vendor names are reported as given. No vendor paid to appear or was able to edit this page.