Google Ads Now Ignores Offline Conversions Uploaded After 7 Days: A D2C Fix

When your best conversions arrive too late to matter

Picture a fairly ordinary week for an Indian D2C brand. Someone clicks a Search ad on Monday, lands on your Shopify store, and places a cash-on-delivery order. Your team confirms it by phone on Wednesday, the courier picks it up Friday, and your operations person exports the confirmed orders and uploads them to Google Ads the following Monday — a tidy weekly ritual. By the time that conversion reaches Google, nine or ten days have passed since the click turned into an order.

For years that was fine. The conversion showed up, your cost-per-acquisition looked right, and Smart Bidding carried on. That quiet arrangement has changed. Google's attribution modelling now bypasses offline conversions uploaded more than 7 days after the event took place — which means a lot of D2C brands are feeding their bidding algorithm a shrinking, skewed slice of their real sales without realising it.

What Google actually documented

This didn't arrive as a splashy announcement. It surfaced inside Google's own Help Center documentation on attribution reports, spotted publicly in late August 2026. The added section states plainly that an offline conversion uploaded more than seven days after the initial event took place is bypassed by the attribution modelling calculations.

There's no separate press release and no loud in-product banner, which is exactly why it's easy to miss. But it's Google's documented behaviour, not a rumour or a test — it describes how the attribution engine treats late uploads right now.

The logic behind it isn't hard to guess. Attribution modelling works best when the signal is fresh, because a conversion that arrives ten days after a click is harder to tie cleanly to the journey that produced it. Whatever the reasoning, the practical effect is what you have to plan around: freshness is now a hard requirement, not a nice-to-have, and the penalty for missing it is silent.

Reporting still shows them — bidding doesn't

Here's the part that trips people up. Your standard conversion columns keep recording those late uploads. Open your campaign report and the totals look complete. What quietly stops happening is the modelling: the "Search funnels" attribution engine that powers your data-driven attribution and, by extension, Smart Bidding no longer factors in conversions that landed past the window.

So you get the worst kind of problem — one that's invisible in the dashboard everyone glances at, while doing real damage underneath. Your reports say you're fine. Your bidding says otherwise, in ways you only feel weeks later as efficiency drifts.

Why this hits Indian D2C harder than most

Plenty of the offline-conversion advice floating around assumes a clean B2B pipeline where a sale closes and gets logged the same day. Indian D2C rarely works like that, and three habits push uploads past 7 days almost by default.

COD confirmation lag

COD is still a huge share of orders for most Indian D2C brands, and a COD order isn't really "done" the moment it's placed. Teams verify by call or WhatsApp, screen for likely RTO, and only then treat it as a real conversion worth uploading. That verification gap alone can eat three to five days. Add a weekend and you're already brushing the cutoff.

Call-centre and manual sales

If part of your revenue closes over the phone — higher-ticket products, repeat buyers, wholesale-ish orders — those conversions live in a spreadsheet or a call log before anyone imports them. Manual handling means manual delay, and manual delay is unpredictable.

CRM and marketplace reconciliation

Brands that sell across their own site plus quick commerce or marketplaces often wait to reconcile everything before uploading a clean file. Reconciliation is sensible for accounting. For attribution, it's poison, because it holds good conversions hostage until the slowest data source catches up.

The models that quietly stop learning

It helps to be precise about what breaks and what doesn't, because the change is narrow but pointed. Here's the practical split.

The takeaway from that table is simple: anything that learns from your data is affected, and anything that merely counts your data isn't. Since bidding is the thing that spends your money, the affected column is the one that costs you.

The 7-day fix: it's cadence, not code

The good news is you don't need a new tracking stack. You need to move conversions into Google faster and more reliably. Three moves cover most of it.

Audit your current upload lag first

Before changing anything, measure the gap between when conversions happen and when you upload them. Pull your last month of offline imports, compare the conversion timestamp against the upload timestamp, and find your average and your worst case. If the average is over 7 days, your bidding has been optimising on a fraction of reality and your reported CPA has been fiction for a while. This one number tells you how urgent the rest of this is.

While you're in there, look at the spread, not just the average. A brand that uploads most orders in two days but dumps a slow tail of COD confirmations at day twelve has a targeted problem worth fixing, whereas a flat weekly batch is a systemic one. The shape of the delay tells you whether you're patching a process or rebuilding it.

Automate the upload — stop batching it weekly

A human exporting a file every Monday is the single biggest source of lag. Shift to a daily or even hourly automated upload straight from your source of truth, whether that's Shopify, your CRM, or a call-centre sheet. If you'd rather not wire up a full API pipeline, a scheduled routine that pushes new confirmed orders each day does the job, and it's exactly the kind of repetitive, deadline-sensitive task that our Google Ads scripts work is built to automate. The goal is boring reliability: conversions leave your system the same day they're confirmed, every day, without anyone remembering to do it.

Handle COD without missing the window

COD is the tricky one because you genuinely can't confirm every order same-day. Two workable approaches: upload the order at the point of confirmation (call or WhatsApp verification) rather than waiting for delivery, since confirmation usually lands well inside 7 days even when delivery doesn't; or, if you must wait for delivery, tighten your verification SLA so confirmed orders never sit un-uploaded. Either way, timestamp the upload with the real order time, not the upload time — Google measures the clock from when the event happened.

What to check this week

You don't need a project for this. You need an hour and a willingness to look at an ugly number.

Start by confirming your offline conversion import is even active and mapped to the right conversion action. Then measure the upload lag as described above. If it's past a week, change the cadence before you touch anything else in the account — no bid strategy tweak matters if the data feeding it is stale. Finally, if you rely on Enhanced Conversions for Leads, remember it improves matching but does nothing about the 7-day cutoff; it's a complement to fast uploads, not a substitute.

The bottom line

None of this makes offline conversions less valuable — they're often your most profitable, highest-intent sales. It just means Google has drawn a firm line around how fresh that data has to be to shape your bidding. Reports will keep flattering you regardless, which is precisely why this is worth acting on now rather than after a quarter of quietly degraded performance.

Check your upload lag. If confirmed COD and CRM sales are reaching Google inside the window, you've lost nothing. If they're arriving a week or two late, your Smart Bidding has been learning from the wrong picture, and the fix costs you a cadence change, not a rebuild.