Open the live demo →
Feedwatch

Know the moment a bank feed breaks.

A budgeting app for people with accounts at several banks. When a bank connection breaks — and they all break — most apps go quiet, so your budget stops being true and you find out weeks later. Feedwatch shows the status of every account up front, and reconnects through a second provider before it gives up.

feedwatch — connection health
Connection health
4 accounts
ChaseChecking ··4471Synced4 min ago
American ExpressPlatinum ··1002Stale6 days
FidelityRoth IRA ··8830Reconnectingattempt 2 of 3
AllySavings ··2214Synced11 min ago

What it does

Shows the truth

Every account with its real connection status and the time of the last successful update — on the first screen, not buried in settings.

Reconnects itself

If one provider can't reach your bank, it tries another before telling you anything is wrong. Most apps only ever try one.

Remembers your rules

Categorise a merchant once and it stays that way. Automatic guessing can be switched off completely.

Why we're building it

The numbers below are about the budgeting apps people use today — not about Feedwatch. We read 4,179 one-to-three-star reviews of budgeting apps written by people who were already paying to work out what actually goes wrong before designing anything.

113 of 4,179 named the bank feed

113

Reviews complaining specifically about a broken bank connection — the highest concentration of any niche measured.

10.1%
of complaints are about billing

10.1%

Billing and cancellation disputes. The highest rate in the study, because the subscription keeps charging while the feed is dead.

1–3★ share, by app

64%

Of Quicken's App Store reviews are one to three stars. YNAB sits at 47%, PocketGuard at 56%.

8 of 30 days with a failed feed

4,179

Low-star reviews read across the budgeting category to produce these figures.

Aggregator

Every product in this category reaches your bank through Plaid or MX. When that link fails the vendor has no direct remedy, and the subcontractor is invisible to the person paying.

Support

Reviews describe weeks of silence while a feed is down. One user logged twenty-two days without a Chase connection and cancelled rather than wait longer.

Categorisation

Auto-detection that overrides the human every month is the single most-upvoted complaint in the niche — 111 votes on one review alone.

How Feedwatch works

1

Connect

More than one aggregator behind every institution. When one provider can't reach your bank we try another before telling you it's broken — because renting from a single subcontractor is why every incumbent fails the same way on the same morning.

feedwatch — provider failover
Fidelity · 09:14
auto
Plaidtimed out — 30s1st
MXconnected — 1.9sused
Resultbalance current, no gap✓
2

Verify

Connection health on the front page, not buried three screens deep. Per-account status with the real timestamp of the last successful pull, so a stale feed is something you learn from us rather than from a budget that quietly stopped being true.

feedwatch — feed uptime
Feed uptime
30 days
30 days ago98.2% uptoday
3

Classify

Correct a category once and it stays corrected. Auto-detection you can switch off entirely. The rules are yours, and the software is not permitted to overrule them every statement cycle.

feedwatch — categorisation rules
Your rules
2 locked
Locked
SQ *COFFEE ROASTERS → Coffee
Applied to 34 transactions. Never re-guessed.
Locked
AMZN Mktp US*2K4 → Household
Applied to 112 transactions. Never re-guessed.
Auto-detect subscriptionsOff. Your rules win.

What people actually say

This is the highest concentration of bank-connection complaints of any niche measured — and the highest rate of billing anger, because these users are paying while it's broken.

“All my Chase accounts have lost connection for the last 22 days. I have emailed support several times and no one will help. Plaid is apparently to blame but no solution.”

Monarch · 1★ · 76 votes

“The automatic transaction import is flawed. It stops syncing frequently. I contacted support three times. It works after each contact, but then breaks again later.”

YNAB · 2★ · 79 votes

“There is no way to stop their auto subscription detection, so my transactions are constantly getting messed up on what they are assigned to.”

Rocket Money · 1★ · 111 votes

Read the full research, including method and limitations →

Feedwatch is built for:

Multi-bank households

Accounts spread across four institutions, where one broken feed quietly invalidates the entire budget.

Ex-Mint migrants

People who already moved once and will not tolerate a second migration caused by the same failure.

Paying through outages

Billing disputes run 10.1% here because subscriptions keep charging straight through the breakage.

Deliberate categorisers

People who classify on purpose and expect the software to remember what they decided.

Pricing

Not chargeable yet — the product is in development. This is what it will cost, published now so nobody is surprised later.

Free

$0

For a single account, or for trying the method before trusting it.

  • 2 connected accounts
  • Connection health and uptime history
  • Manual and CSV import
  • No second-provider failover
  • No locked rules
Try the demo
Most people

Standard

$8 / month

Or $72 a year — two months free. Everything the product does.

  • Unlimited connected accounts
  • Multi-provider failover on every institution
  • Full 30-day connection history per account
  • Locked categorisation rules, auto-detect off
  • Export everything, any time, no ceremony
Join the waitlist
One-time

Founding

$149 once

Three years of Standard included, then your renewal is locked at today's annual price for as long as you keep it.

  • Everything in Standard
  • Three years included from launch day
  • Renewal locked at $72/year, in writing
  • Direct line to us on what gets built next
Register interest

Why there is no lifetime plan

Every connected bank account costs us money every month, for as long as it stays connected. A true lifetime price would eventually have to be withdrawn, throttled, or quietly degraded — which is precisely the complaint that produced this product in the first place. So the Founding plan locks a renewal price rather than pretending the running cost disappears. That promise is one we can actually keep.

What's actually built

There is a running demo at /app — synthetic accounts, no sign-up, nothing to install. Here is exactly where the line sits between what works and what doesn't.

Working in the demo

  • Connection status derived from sync attempts that really ran — not a stored flag
  • Automatic failover to a second provider, with the trace of both attempts
  • 30-day feed uptime computed from that same history
  • Category rules that rewrite past transactions and report how many changed
  • Transaction list, filtering, and per-account detail

Not built yet

  • Real bank connections — the demo simulates the institution
  • Accounts, sign-in, and anything touching real money
  • Budgets, goals and forecasting
  • Mobile apps — the demo is web only
  • Shared household access

Open the demo →    See each screen explained →