Skip to main content
Metrics type: Key Metrics • Category: Marketplace
Triggered when /outbox/listings has no confirmed ack in 48h, feed is blind, all listing changes stuck.

​
At a glance

P1 alert. Fires when the AbeBooks outbound listings feed (/outbox/listings) has not received a confirmed ack from AbeBooks in 48 hours. While the alert is active, every listing edit (price change, stock change, condition update, new listing, delete) is stuck in a queue with no path to AbeBooks. The catalogue is frozen on whatever state the last successful cycle left it in.

​
Calculation

Calculated automatically from your AbeBooks data. See the At a glance summary above for what the metric tracks and the worked example below for a typical reading.

​
Worked example

A UK independent bookseller, 38,400 active AbeBooks listings, daily outbound feed cycle. Incident timeline 28 Apr 26 to 01 May 26. Cumulative impact during the 56-hour outage: Six things to notice that are specific to AbeBooks and the book-trade FTPS feed model:
  1. 48 hours is the standard threshold but not the only one to use. A bookseller running a daily feed cycle should alert at 48h; a bookseller running a 4-hourly cycle should alert at 12h or 24h. Tune the threshold to your actual feed cadence; default catches the daily-cycle audience.
  2. The “queued, no ack” signature is the most common version of feed failure. AbeBooks’s FTPS server doesn’t send error responses to most failure modes; it just drops the connection or stops processing. Detection is by ABSENCE of ack, not by error message. This is why the threshold is time-based rather than error-count-based.
  3. TLS certificate rotation is the #1 cause and the most preventable. Merchants using their own outbound FTPS client should subscribe to AbeBooks’s developer mailing list for advance notice of certificate rotations (typically 30 days advance notice for planned rotations, 0 days for emergency rotations). Vortex IQ’s connector handles cert pinning automatically, but custom feed clients often miss this.
  4. Stockout cancellation cascade is the second-order damage. During the 56h outage, 412 books sold on other channels were still listed as available on AbeBooks. Of the 14 that buyers attempted to purchase on AbeBooks, every one resulted in a stockout cancellation, which counts toward the 3% cancellation threshold. A 56h outage during a busy week can wipe out 30 days of careful cancellation-rate management.
  5. The 30 to 90 day reputational tail. AbeBooks’s seller-standing tile averages cancellation rate over 30 days. A 14-cancellation cluster during this outage will sit in the 30D average for the next 30 days; expect the seller-standing tile to show “at risk” until early Jun 26.
  6. The 9,840 queued repricer updates are interesting. Most of those are routine 1 to 5 pence pricing nudges that don’t matter individually, but in aggregate they represent the catalogue tracking the live market. Letting them queue and fire all at once after the outage causes a one-day price-change spike that some buyers notice; a phased re-upload (1,000 listings/hour) is gentler.

​
Sibling cards merchants should reference together

Feed cycle failure is the P1 operational alert. Pair with these to triage and recover:

​
Reconciling against the vendor’s own dashboard

Where to look in the AbeBooks seller dashboard: AbeBooks’s seller dashboard does NOT publish a feed-health view; the merchant typically only learns of a feed failure via downstream symptoms (drop in revenue, spike in cancellations). Three indirect views help triangulate during an active alert:
  1. My AbeBooks → Inventory → Upload History. Shows the last N upload attempts AbeBooks recorded; if the most recent entry is >48h old, AbeBooks confirms it has not received your files.
  2. My AbeBooks → Inventory → Manage. Spot-check any listing you know you’ve recently edited; if the AbeBooks UI shows the OLD price/stock/condition, the change hasn’t propagated.
  3. AbeBooks Developer Status Page (when available; intermittent service). Lists known platform-side outages.
Why our alert may legitimately differ from AbeBooks’s own observations: Cross-connector reconciliation: Feed cycle failure is fundamentally per-marketplace; cross-connector reconciliation surfaces shared root causes (merchant-side network, DNS, time service).

​
Known limitations / merchant FAQs

The alert just fired. What do I do in the next 30 minutes? Three checks, in order: (1) Test FTPS auth manually. From a known-good machine, attempt FTPS login to AbeBooks’s host with the connector’s stored credentials. If it fails, you have a credential or TLS-cert issue, fix that. (2) Check the connector’s outbox staging directory. If files have queued for >48h, processing has been blocked, validate disk space, file permissions, and any rotation/cleanup cron jobs. (3) Open AbeBooks’s developer status page. If AbeBooks is in the middle of a maintenance window, the alert will auto-clear when their service resumes; no action needed but document for compliance. My alert just cleared, but my cancellation rate is still climbing. Why? The alert clears when the next ack is received; the operational damage from the outage takes 7 to 30 days to age out. During the outage, books that sold elsewhere stayed live on AbeBooks; buyers who attempted to purchase those got stockout cancellations. Those cancellations stay in the rolling 30D average for the next 30 days. Watch Cancellation Rate for the recovery. Multi-marketplace, can a single network outage trip both AbeBooks and Alibris feed alerts? Yes, if the cause is at the merchant data centre (DNS, ISP, firewall rule). If the cause is at the marketplace side (AbeBooks-only outage, Alibris-only outage), only one alert fires. The first triage step when both alerts fire together is “is my outbound network healthy?”, not “is AbeBooks up?”. Listing-quality / Buy Box impact, what’s the second-order damage? The cancellation cascade described in the worked example is the primary second-order effect. Beyond that: AbeBooks’s Listing Quality Score sees no positive signals during the outage (no new listings, no condition updates, no fresh imagery), and the score decays slightly. Recovery to baseline takes 2 to 4 weeks of normal feed operation. Inventory-sync lag during the outage, what’s the right behaviour? Two strategies: (a) Conservative, when you detect feed failure, manually mark all AbeBooks listings as “out of stock” via the AbeBooks website upload to prevent stockout cancellations. Pros: no cancellations during outage. Cons: zero AbeBooks revenue during outage, and re-enabling 38,000 listings post-outage causes a temporary AbeBooks ranking suppression. (b) Permissive, leave listings live, accept the stockout cancellations, recover post-outage. Pros: some AbeBooks revenue continues. Cons: cancellation-rate damage. Most booksellers under 1,000 daily orders use (b); larger sellers use (a). ISBN match quality, can it cause feed failures? Indirectly. Malformed ISBN-10 / ISBN-13 in your outbound file can cause AbeBooks to silently reject the entire file (not just the malformed rows) on some legacy AbeBooks parsers. The signature is “card alerts after a recent inventory-data refresh”. Open Top Upload Error Types immediately after recovery; ISBN-format violations dominate when this pattern is the cause. Why doesn’t AbeBooks just send an email when my feed breaks? AbeBooks does send email warnings on persistent failures, but the email cadence is once-weekly batch and the threshold is 7+ days, far slower than the 48h alert this card provides. Treat AbeBooks’s own emails as a backstop, not a primary alert mechanism. Rare books vs commodity books, does the outage hurt them differently? Yes. Commodity books have high volume, so each frozen-feed day produces multiple stockout cancellations on the popular ISBNs. Rare books have low volume; a 56h outage may produce ZERO rare-book cancellations because no rare-book buyer happened to attempt purchase during the window. The outage’s reputational impact concentrates on commodity stock. When does the alert turn off vs when do I act? The alert auto-clears when the next ack is received. But “act” until you’ve also: (1) drained the outbox queue (re-uploaded queued files); (2) confirmed all listings are in their intended state via spot-check on AbeBooks’s UI; (3) reviewed the cancellation cluster from the outage and proactively contacted any affected buyers with a discount code or apology to limit feedback damage. Full recovery is 2 to 4 weeks; the alert clearing is just the first 5%.

​
Tracked live in Vortex IQ Nerve Centre

Feed Cycle Failure is one of hundreds of KPI pulses Vortex IQ tracks across AbeBooks and 70+ other ecommerce connectors. Nerve Centre runs the detection layer; Vortex Mind investigates the cause when something moves; Ask Viq lets you interrogate any number in plain English. Start for free or book a demo to see this metric running on your own data.