← Glossary / Denial of Inventory Attack

What is Denial of Inventory Attack?

A Denial of Inventory Attack (or cart hoarding) is a malicious, state-changing bot operation where automated scripts add high-demand items to shopping carts to trigger a platform's temporary inventory reservation system. By holding stock hostage until the session expires and immediately re-adding it, attackers artificially deplete available inventory, preventing legitimate customers from purchasing while the bot operator resells the items elsewhere.

Stateful BotsE-commerceCart HoardingScalpingWAF Rules
// 02 — definitions

Locking up
the shelves.

The mechanics of how stateful bots weaponize cart expiration timers to artificially deplete stock without ever completing a purchase.

Ask a DataFlirt engineer →

TL;DR

Unlike web scraping, which passively reads data, a denial of inventory attack actively alters server state. It exploits the 10-to-15 minute "cart lock" most e-commerce platforms use to guarantee availability during checkout. A distributed botnet can lock up thousands of SKUs in seconds, causing massive revenue loss and customer frustration.

01Definition & structure
A Denial of Inventory Attack occurs when automated scripts exploit an e-commerce platform's cart reservation system. When a user adds an item to a cart, the backend database typically places a temporary lock on that SKU to ensure it isn't sold to someone else during the checkout process. Attackers use distributed botnets to add thousands of items to fake carts simultaneously, dropping the available public inventory to zero without ever spending a dime.
02How it works in practice
The attacker targets a high-demand product (like a sneaker drop, a graphics card, or concert tickets). They deploy a fleet of headless browsers routed through residential proxies. Each bot creates a guest session, sends a POST request to the cart endpoint, and holds the item. If the cart expires in 15 minutes, the bot is programmed to drop the item at 14:59 and immediately re-add it at 15:00, creating a sustained denial of service for legitimate buyers.
03The economic motive
Inventory attacks are rarely done just for vandalism. They are usually the first phase of a scalping operation. By locking up the primary market's inventory, the attacker creates artificial scarcity. They then list the items on secondary markets (like eBay or StockX) at a massive markup. Once a buyer purchases the item on the secondary market at the inflated price, the bot operator finally completes the checkout flow on the primary site, pocketing the arbitrage margin.
04How DataFlirt handles it
We don't. DataFlirt is a data infrastructure company, not a botting service. We strictly enforce read-only operations across our entire fleet. Our proxy gateways are configured to drop state-mutating HTTP methods, and our crawler configurations actively exclude cart, checkout, and account-creation paths. We extract the data you need to compete, but we never touch the inventory.
05Did you know?
Some sophisticated inventory attacks don't even target the cart endpoint. They target the payment gateway integration. By initiating a payment intent and stalling the 3D Secure or third-party redirect flow, bots can trigger a much deeper, longer-lasting database lock on the inventory than a standard cart addition, sometimes holding stock for hours instead of minutes.
// 03 — the attack model

Calculating
inventory lockup.

The severity of an inventory attack depends entirely on the target's cart reservation policy and the botnet's concurrency. If the lock time exceeds the bot's re-add latency, the inventory is permanently denied.

Lockup capacity = L = bot_concurrency × max_cart_qty
Total items a botnet can hold simultaneously across all active sessions. E-commerce threat modeling
Sustained denial condition = Texpiry > Tre-add
If the bot can drop and re-add the item faster than a human can checkout, the item remains locked. Bot operator tactics
Business impact = Cost = locked_items × conversion_rate × margin
The actual lost revenue per hour of sustained attack. Retail security metrics
// 04 — the WAF trace

A cart hoarding
sequence.

A trace from an e-commerce edge node detecting a distributed inventory attack targeting a limited-edition sneaker drop. The bots are cycling sessions to maintain the lock.

POST /cart/addSession hoardingRate limit bypass
edge.dataflirt.io — live
CAPTURED
// inbound traffic spike detected
target.sku: "SNK-882-BLU-10"
inventory.available: 450

// distributed POST requests
req.method: POST /api/v1/cart/add
req.ip: 104.28.x.x // residential proxy
session.id: "sess_9a8b7c..."
cart.status: 200 OK // item reserved for 15m

// 4 seconds later
inventory.available: 0 // stock fully depleted
checkout.rate: 0.0% // no actual purchases

// WAF intervention triggered
rule.match: "cart_add_without_checkout_flow"
action: invalidate_sessions
inventory.restored: 450
mitigation: enforce_captcha_on_add
// 05 — attack vectors

How bots hold
the inventory.

The operational tactics used by scalper bots to maximize lockup time while evading standard rate limits and IP bans.

TARGET SECTORS ·  ·  ·    Retail, Ticketing, Travel
AVG LOCK TIME ·  ·  ·  ·  10–15 minutes
ATTACK INTENT ·  ·  ·  ·  Resale / Sabotage
01

Distributed guest carts

High concurrency · Creating thousands of anonymous sessions to bypass account limits
02

Session renewal looping

Sustained denial · Dropping and re-adding items milliseconds before the cart expires
03

Max-quantity abuse

Efficiency · Adding the maximum allowed SKU count per cart to minimize required sessions
04

Checkout step stalling

Deep lock · Progressing to the payment step to trigger a harder, longer inventory lock
05

Competitor sabotage

Malicious · Locking up a rival's inventory to drive traffic to their own site
// 06 — the boundary

We read data,

we never change state.

There is a hard, non-negotiable boundary between web scraping and inventory attacks: state mutation. DataFlirt operates strictly read-only pipelines. We extract pricing, catalog, and availability data using HTTP GET requests. We never invoke a POST request to a cart endpoint, we never create accounts to hoard items, and our infrastructure is physically restricted from executing checkout flows. Denial of inventory is a malicious attack; data extraction is competitive intelligence. We build the latter.

DataFlirt pipeline constraints

How we enforce read-only operations across our scraping fleet.

http.methods.allowed GET, HEAD, OPTIONSenforced
http.methods.blocked POST, PUT, DELETEdropped at proxy
cart.endpoints /cart/*, /checkout/*blacklisted
session.persistence statelessno cookies held
inventory.impact 0 items lockedverified
pipeline.intent data extraction only

Stay ahead of the pipeline

Data engineering
intel, weekly.

Anti-bot shifts, scraping infrastructure updates, dataset delivery patterns, and business outcomes from our pipelines. Short, technical, no fluff.

By submitting, you acknowledge our Privacy Notice.

// 07 — FAQ

Common
questions.

Common questions about the mechanics of inventory hoarding, how it differs from scraping, and how platforms defend against it.

Ask us directly →
What is the difference between web scraping and a denial of inventory attack? +
Web scraping is passive and read-only; it fetches HTML or JSON to extract data like prices and stock levels without altering the server's state. A denial of inventory attack is active and state-mutating; it uses POST requests to add items to carts, triggering backend database locks that temporarily remove the item from available public stock.
Why do e-commerce sites allow carts to lock inventory in the first place? +
It's a UX requirement. If a customer adds an item to their cart and proceeds to enter their shipping and payment details, the platform must guarantee the item won't sell out before they hit "Confirm." This temporary reservation (usually 10–15 minutes) is exactly what the bots exploit.
Is a denial of inventory attack illegal? +
In many jurisdictions, yes. Unlike scraping public data, which is generally protected, intentionally disrupting a business's operations by exhausting server resources or locking up inventory often violates anti-hacking laws (like the CFAA in the US or the Computer Misuse Act in the UK) and constitutes tortious interference.
How do WAFs detect cart hoarding bots? +
Modern WAFs look for behavioral anomalies: high volumes of POST /cart/add requests without corresponding progression to the checkout URL, sessions that expire and immediately re-add the same SKU, and massive spikes in guest cart creation from residential proxy IPs.
Can CAPTCHAs stop inventory attacks? +
They can mitigate them if placed correctly. Putting a CAPTCHA on the "Add to Cart" button stops naive bots, but introduces massive friction for real users. Attackers also use CAPTCHA-solving farms to bypass this. The better defense is backend logic: limiting cart quantities per IP/device fingerprint and shortening lock times for high-demand drops.
How does DataFlirt ensure it doesn't accidentally trigger inventory locks? +
Our proxy and worker infrastructure is configured to drop any state-mutating HTTP methods (POST, PUT, DELETE) by default. We only execute GET requests. Furthermore, our URL discovery queues explicitly blacklist common cart and checkout path patterns. We physically cannot add an item to a cart.
$ dataflirt scope --new-project --target=denial-of-inventory-attack READY

Tell us what
to extract.
We do the rest.

20-minute scoping call. Pilot dataset within the week. Production within two. Whether you need a one-off catalogue dump or a continuous feed across millions of records — we scope, build, and operate the pipeline.

Start a pipeline → View pricing
hello@dataflirt.com  ·  Bengaluru  ·  IST  ·  typical reply < 4h
// 08 — keep reading

Related glossary terms

Scraping Security

Scalping Bot (Retail)

The specific type of bot operator that utilizes inventory denial to secure high-demand goods.

Read →
Scraping Security

Ticket Scalping Bot

How inventory hoarding is applied to the live events industry to corner the secondary market.

Read →
Web Scraping

Stateful Scraping

Why maintaining session state is sometimes necessary for data, but dangerous if it mutates backend records.

Read →
Network Layer

POST Request

The HTTP method that separates passive data extraction from active state mutation.

Read →