Most auctioneers put off switching software for one reason: the last time they migrated anything, it was a nightmare. A weekend of re-typing, a bidder list that half-imported, an invoice run that went out with the wrong buyer’s premium. Nobody wants to relive that during a live sale with consignors watching.
Here’s the good news, and it’s structural: an auction house is one of the lower-risk businesses to re-platform. Your inventory flows through — a lot you catalog today is gone in two weeks, so there’s almost no “back catalog” you’re forced to drag across. The things that actually have to move are small, portable, and mostly live in spreadsheets already: your bidder list, your consignor list, and your historical sales numbers. If you can export those cleanly, most of the risk is already behind you.
This is a checklist for doing it without drama. It assumes you’re moving your cataloging, catalog-publishing, and settlement workflow — not your bidder audience. If you run on a marketplace like HiBid or Proxibid for reach, the honest move is to keep paying for that reach while you switch the workflow underneath it. You don’t have to choose. More on that below.
The one rule that de-risks everything: never cut over cold
The single mistake that turns a migration into a disaster is the big-bang cutover — canceling the old system on Friday and running your next sale on the new one Monday. Don’t. Run both in parallel through exactly one sale. The old system stays fully live and paid-for; the new one gets a real, low-stakes auction to prove itself on. If it fails, you’ve lost nothing but a little duplicate effort. If it works, you cut over on your own timeline, with evidence.
Everything below is organized around that: get your data portable first, pilot on one sale, measure two numbers, then decide.
1. Before you commit: run the export test
The most important thing to check about your new vendor isn’t a feature — it’s whether you can get back out. And the most important thing to check about your old vendor is whether you can get your data out of it at all. Do this before you sign anything.
Confirm each of these before you commit:
- ☐ Can you export your bidder list from the old system as a CSV, yourself, today? Names, emails, phones, opt-in status. If the answer is “email support and wait,” or “we don’t offer exports,” that’s not a migration problem — that’s a reason you should have left already.
- ☐ Can you export your consignor / seller list the same way?
- ☐ Can you pull your historical sales results — hammer prices by lot, by sale — for at least the last year? You want this for pricing comps, not for import, but you want it.
- ☐ Does the new system let you export all of the above yourself, no ticket, no fee? Ask them to show you the button on a demo, not describe it. A vendor that makes it easy to leave is telling you something true about how they treat your data.
If the marketplace or platform you’re on today owns the bidder relationship — the registrations are on their domain, the card-on-file is in their vault, the invoices come from their from-line — understand that you may not be able to extract everything. You can usually get the list of humans who bought from you; you generally cannot extract their stored payment methods, and you shouldn’t expect to. Plan for bidders to re-register (section 5).
2. Get your data out
Do the export while your old contract is still active and your account is in good standing. People forget this and try to pull their data after they’ve given notice, when access is already winding down. Order of operations:
- Export the bidder list to CSV. Open it. Actually look at it. Confirm the columns you care about are populated — a “phone” column that’s 80% blank is worth knowing about now.
- Export the consignor list to CSV. Same check.
- Export historical sales results. Keep this as your own archive regardless of which system wins. Your own sold-price history is the best pricing comp you own, and no platform should hold it hostage.
- Note your opt-in / consent status per bidder. This is the field people drop, and it’s the one that gets you in trouble. A bidder who opted into your emails on the old platform did not necessarily opt into the new one. Carry the consent flag across; don’t assume it.
3. Bring it in (and be realistic about what “import” means)
Here’s where honesty matters more than a marketing promise. A clean import of a bidder or consignor CSV is a solved problem — it’s a list of people with contact fields, and any competent system will take it. That part should be genuinely easy.
What is not a solved problem — from any vendor — is faithfully migrating years of past catalog structure, lot-level media, and settlement records between two systems that model auctions differently. The reassuring truth is you almost never need to. You’re not selling last spring’s lots again. Bring across the two things that compound — the people and the price history — and let the old catalog stay archived where it is. If you ever need to look up what a lot sold for in 2024, you have the CSV from section 2.
So the honest import checklist is short:
- ☐ Bidder list CSV imported and spot-checked (pick ten names, confirm they came in with phone + email + consent intact).
- ☐ Consignor list imported.
- ☐ Historical sales kept as your own archive file — not imported, just kept.
- ☐ A decision, written down, on which past records you are deliberately not migrating, so nobody goes looking for them in a panic mid-sale.
4. Contract timing: don’t give notice until the pilot is done
This is where money leaks. The instinct is to cancel the old thing the moment you sign the new one, to stop paying twice. Resist it. You will pay for one month of overlap, and that month is the cheapest insurance you’ll ever buy.
Work the calendar backward:
- Find your old contract’s renewal / notice date. Annual auction contracts love to auto-renew with a 30- or 60-day notice window. Know yours before you plan anything.
- Schedule the pilot sale (section 6) so it finishes with room to spare before that notice deadline. You want the pilot done, settled, and evaluated while you still have the option to stay put.
- Prefer month-to-month on the new system for the trial period. Look for a real trial or money-back window rather than an annual prepay. If a vendor won’t let you test on one sale without a year’s commitment, that tells you how confident they are.
- Only give notice on the old system after the pilot clears your go/no-go bar. Not before.
For what it’s worth: Hammerwerks is month-to-month ($99/mo + $49/auction), with a 30-day money-back window, and founding customers get the first auction free with $0 setup — so the pilot sale itself adds nothing on top of the monthly fee, and the money-back window covers even that. Whatever vendor you evaluate, insist on terms that let you back out cheaply. The ones worth using offer them.
5. Bidder re-onboarding: the one piece of real friction
If any part of this migration is going to sting, it’s this one, so let’s be straight about it. Your bidders may have to re-register on the new system — a new account, and (once, at first purchase) a card on file. You generally cannot port a stored payment method from one platform to another, and you wouldn’t want a vendor who claimed they could. So there will be a first sale where returning bidders hit a signup step they didn’t have to before. Plan for it instead of being surprised by it.
What makes it manageable:
- ☐ Put the new catalog on your own domain, under your own name. A bidder who has bought from you for years will re-register at your auction house without much thought. They would hesitate at a random new brand. This is the single biggest lever on re-registration rate, and it’s why white-label matters here beyond vanity.
- ☐ Warm them up before the pilot sale. One plain email from you — from your from-line, not a platform’s — a week ahead: “We’ve moved our online bidding to our own site. Same auctions, quick one-time signup, here’s the link.” No hype. Your regulars will do it because they trust you, which is the whole asset.
- ☐ Keep your marketplace channel running in parallel for the pilot. If you list on HiBid or similar for reach, keep listing there yourself. Bidders who only ever knew you through the marketplace can keep bidding the way they always have while your own channel finds its feet. You are not asking anyone to switch cold either.
- ☐ Expect the re-registration rate on sale one to be partial, and rising after. Your most loyal bidders move first. The occasional ones migrate over the next few sales as they see you keep showing up on your own site. This is normal. Don’t judge the whole decision on the first night’s signup count.
Set expectations honestly with yourself here: no software makes a switch invisible to your bidders. What good software does is make the one-time re-registration fast, put it under your brand so people trust it, and make sure that once a bidder is on your list, they’re on your list — exportable, yours, and never marketed to on the vendor’s behalf.
6. Pilot on exactly one sale
Pick the right sale to pilot on. Not your biggest estate of the year, and not a throwaway with twelve lots either — you want something representative enough that the numbers mean something, small enough that a hiccup doesn’t cost you a marquee consignor.
The pilot protocol:
- Catalog the whole pilot sale in the new system, start to finish, exactly the way you’d run it for real. If the new system drafts lots from photos, use it as intended: shoot the photos, let it draft, then review and approve every lot yourself. Don’t half-use it and then judge it.
- Keep the old system available but idle as a fallback for that sale — still paid, still there, so a bad night has an exit.
- Run the sale. Keep listing on your marketplace for reach the way you always do, and take bids on the new white-label catalog alongside it.
- Settle it fully in the new system. Invoice, charge, seller statements — all the way through. Settlement is where a lot of software quietly falls apart, so it’s the part you most need to see finished, not just started.
- Debrief the same week, while it’s fresh, against the numbers in the next section.
7. What to measure
Don’t evaluate on vibes. A demo always feels good; a switch has to pay for itself. Measure a small number of things you can actually compare against your old workflow.
The two numbers that decide it
- ☐ Catalog time per lot. Total hands-on time to get the sale cataloged and approved, divided by lot count. This is the number the whole workflow lives or dies on. Compare it honestly to what the same sale took you before. Photo-to-draft cataloging should move this — but count your review and correction time in it, because approving every lot is a real step, not a formality. A tool that drafts fast but forces you to rewrite every line hasn’t saved you anything.
- ☐ Settlement time. From last-lot-hammer to invoices out, payments captured, and seller statements drafted. If the old way was a QuickBooks export and an evening of manual invoicing, and the new way is automatic, this is where a flat monthly fee quietly justifies itself. Time it on both.
Worth tracking too
- ☐ Bidder re-registration rate on the pilot sale — and the trend over the next one or two sales, which matters more than the first night alone.
- ☐ Draft accuracy you can trust. Roughly what fraction of AI-drafted lots needed only light edits vs. a full rewrite. This tells you whether the time savings is real or a mirage.
- ☐ Settlement correctness. Did buyer’s premium, taxes, and seller commissions compute right, every lot? One wrong invoice erodes more trust than ten slow ones.
- ☐ Did anything require a support ticket, and how long did it take to resolve? You’re not just buying software; you’re buying whoever answers when it breaks.
8. The go / no-go
Set the bar before the pilot, not after, so you’re not grading on a curve you moved. A reasonable bar: catalog time per lot is meaningfully lower, settlement ran clean and faster, and your loyal bidders re-registered without a support fire. Clear it, and give notice on the old system. Miss it, and you’ve spent one month of overlap and one duplicate sale’s effort to learn something real — which is exactly what the overlap was for. Either outcome is a win over guessing.
Hammerwerks was built to be piloted this way. Your bidder list imports from CSV and exports back out whenever you want, no ticket and no fee — you’re never locked in. The catalog runs white-label on your own domain, so re-registration lands under your name, not ours. And it runs alongside the marketplaces you already use for reach — keep those for the audience, since we don’t pretend to match a HiBid-sized bidder pool — so you’re never asking bidders to switch cold. It’s month-to-month with a 30-day money-back window and a free first auction for founding customers, specifically so the pilot sale costs you almost nothing to try. If you want to walk through what migrating your next sale would actually look like, book a 15-minute call or watch the 90-second demo first.