Hubdoc discontinued Auto-Fetch: what firms use now

If you landed here, you probably just found the words “Hubdoc pulls these automatically” in an old workflow doc, or a new hire asked why the client folders stopped filling themselves and nobody remembered the answer. The short version: Hubdoc no longer fetches bank statements. Xero retired Auto-Fetch in early 2022, and it never came back. Hubdoc itself is still around. It takes uploads, extracts data from documents, pushes to Xero, and comes included with a Xero subscription. The part that went out and got the statements for you is gone.

That retirement left a real hole, because statement retrieval was the one job almost nothing else in the document stack was doing. This post covers what Auto-Fetch did, why killing it was a defensible call, where firms actually migrated, and what to check before you commit to a replacement.

What Auto-Fetch actually did

Auto-Fetch was the automated-retrieval half of Hubdoc. Once a client’s banks were connected, recurring PDF statements showed up in the client’s folder on their own, ready for extraction and the push to Xero. For bookkeeping teams, that meant month-end packets that assembled themselves in the background.

The fetching mattered less than the remembering. Nobody had to keep a spreadsheet of which statements were due, nobody had to email the client, and nobody had to log into fourteen bank portals on the third of the month. When Auto-Fetch went away, all of that remembering came back onto someone’s desk.

Why Xero pulled the plug

It is worth being fair here, because the decision made engineering sense. Automated retrieval of that era mostly worked the way a person would: store the login, sign in to the bank’s website, click around, download the PDF. That approach is fragile in ways that compound. Banks redesign login pages without notice. They add multi-factor prompts. They deploy bot detection that cannot tell a fetch tool from an attacker, and frankly should not have to. Every institution is its own small integration, and each one breaks on its own schedule. We wrote up the mechanics in why bank statement connections break if you want the longer version.

The worst part is that these failures are quiet. A fetch that stops fetching does not raise an alarm at the client’s bank. The vendor eats a support queue that never shrinks, and the firm discovers the gaps at close, which is the most expensive possible moment to discover anything.

Xero looked at that treadmill and stepped off. Retiring a feature you cannot run reliably is more honest than leaving it up while it decays. The timing also tracked a bigger shift: bank access was moving from screen scraping toward consented, API-based open banking, including statement endpoints at Plaid that deliver the bank’s official PDF without anyone storing a password. The scraping generation was ending either way.

The five paths firms took after 2022

Straight back to manual downloads

This is where most teams landed, at least at first. Someone on staff logs into each portal, or the client gets a reminder email and does it themselves. It works. It also costs you every single month: chasing, MFA codes texted back and forth, statements for the wrong period, and the occasional screenshot of a balance instead of a PDF. Our own close stalled waiting on statements more months than we care to admit. The full cost comparison lives in manual vs automated statement collection.

Request tools and client portals

Content Snare, ShareFile, Liscio, and similar tools organize the chase: templated requests, automatic reminders, a secure place to drop files. That is genuinely better than email threads. But the model has a ceiling, because the client still does the work every month, and clients get tired. A reminder tool makes the ask politer and more trackable. It does not remove the ask.

LedgerSync covers the most ground

If you want a like-for-like replacement, LedgerSync is the obvious candidate, and in several ways it goes further than Auto-Fetch did. It fetches transactions, statements, and check images for accountants, covers 10,000+ banks through Mastercard Open Banking and MX, and pushes into QuickBooks Online, QuickBooks Desktop, Zoho Books, Accounting CS, or CSV. There is rules-based and AI reconciliation, OCR, mobile client capture, and an open API. Pricing is per active client and drops with volume; a 100-client firm pays roughly $16.35 per client per month as of July 2026, with a 30-day free trial.

The record is not spotless. Reviewers on G2 and Capterra report connection reliability problems since the LS2 rewrite, statements occasionally filed to the wrong account or period, price increases with V2, check images moving behind an add-on fee, and a $34.90 monthly minimum that some say kept billing on inactive accounts. Public reviews also mention that vendor support staff can view fetched documents, which matters if custody is a concern for you. Even so, it is mature, it has years in market, and it does far more than statements. We compare the field in LedgerSync alternatives.

LedgerDocs sells fetching as an add-on

LedgerDocs approaches this from the document-management side. The core product is OCR data extraction, PDF-to-CSV conversion, mobile receipt capture, and a QuickBooks Online integration, with plans running roughly $11 to $67 per month billed annually as of July 2026 and a 14-day trial. Statement fetching is an add-on priced around $3 per fetch, with a $9 monthly minimum that includes three fetches, as of July 2026. For a firm covering a handful of accounts, that is an inexpensive way in, and you get the whole receipts-and-bills lifecycle alongside it. The per-fetch math changes as account counts grow, so run your own numbers before assuming it stays cheap.

StatementFlow, the one we build

Disclosure first: this is our product, so read the next paragraph with that in mind. StatementFlow came out of Scale CPA, a US accounting firm serving tech startups, SaaS, ecommerce, and restaurant clients, and it runs on the firm’s own client book first.

It does one job. Each client connects their bank once through a secure invite link, authenticating at their own bank via Plaid or Mastercard Open Banking, so the firm never sees credentials. The system learns each account’s real statement-posting cycle, fetches the official PDF when the bank publishes it, verifies every download with a SHA-256 hash, and files it into the firm’s own Google Drive, organized by client, year, and month. A coverage board (accounts by months) surfaces gaps, needed reconnects, and known bank security holds at Wells Fargo, Chase, Bank of America, and Citi before close. It also pulls settlement statements from Stripe, PayPal, Square, Shopify, Etsy, eBay, Wise, and the major ecommerce platforms.

What it does not do: no OCR, no GL push, no client portal, US banks only. It is in early access with a waitlist and no public pricing yet. Security details, including token encryption and passkey step-up, are on the security page.

Here is the whole landscape in one place:

Four things to check before you commit

Coverage against your actual client list. Vendor bank counts are marketing numbers. Write down the twenty institutions your clients actually use and ask about each one, including how the tool behaves at banks with aggressive security-hold flows. Also ask how it decides when a statement is due, because posting cycles routinely ignore calendar months, and a tool that checks on the 1st will miss statements that post on the 17th.

Custody of documents and tokens. Where do the PDFs land: the vendor’s cloud, or storage you control? Who at the vendor can open them? What is the export path if you leave? Under the FTC Safeguards Rule, your written security program has to account for service providers holding client financial data, so these are compliance questions, not just preferences.

Verification of what got fetched. Ask whether you receive the bank’s official PDF or a re-rendered copy, whether downloads are integrity-checked, and how the tool confirms a statement was filed to the right account and period. Reviewers of at least one major tool in this category report statements filed to the wrong account or period, so do not treat filing accuracy as a given.

Reconnect handling. Every connection eventually breaks: expired consent, a changed login, a security hold. The differences between tools show up here. How fast do you find out? Is there a view of every account and month so a gap is visible at a glance? And when a reconnect is needed, does the client get a link they can tap, or does someone open a support ticket?

So which one should you pick?

It depends on the job. If you need transactions, check images, and GL pushes alongside statements, start a LedgerSync trial and point it at your ten hardest banks for the full 30 days. If you already live in LedgerDocs or only cover a few accounts, the fetch add-on is cheap enough to just try. If the job is specifically official statements, verified and filed into storage you control, that is the gap we built StatementFlow to fill; it is early access, and you can join the waitlist here. For a wider survey, we maintain a round-up of statement retrieval software and a direct Hubdoc comparison.

Whatever you choose, do not rebuild the old workflow doc around a promise of silent automation. Auto-Fetch taught the industry that lesson once. Fetching breaks, and the tools worth paying for are the ones that tell you when it does.