What do Acumatica users report about the native commerce connectors?

On 20 August 2026 we read the Acumatica Community forum named Retail, subtitled "Retail Edition QnA", at community.acumatica.com/retail-113. The same forum answers to community.acumatica.com/commerce-113. On the day we read it, the forum header reported 648 topics and 3,202 replies.

We read 20 index pages: the forum landing page, plus index2.html through index20.html. Each index page listed 25 threads, giving 500 listing rows. The pagination overlaps in two places. Index page six repeated 20 rows already listed on index page five, and index page eight repeated eight rows already listed on index page seven. We deduplicated on the numeric thread ID at the end of each thread URL, which removed 28 duplicate listings and left 472 unique threads. That is 72.8 per cent of the 648 topics the forum reported, so 176 topics were outside what we read.

Two pairs in the sample look like accidental double-posts by the same person: threads 26182 and 26183 ask the same sandbox question, and threads 24908 and 24909 ask the same branch-mapping question. Each pair has two distinct thread IDs, so we counted them as two threads each rather than silently merging them. That decision inflates the count by two out of 472, and we would rather state it than hide it.

We classified each thread from its title alone. We did not open all 472 threads. We opened a small number to check specific claims, and every thread we opened is linked below with the date it was posted.

Every thread got exactly one category. The classifier is a fixed list of keyword rules applied to the title in a fixed precedence order, first match wins, and the order is: tax, payments, pricing, products and variants, inventory and warehouses, customers and accounts, orders and fulfilment, errors and sync failures, configuration, other. Categories are subject-first, so a thread titled as a pricing error counts as pricing rather than as an error. That is a choice, and it is why the errors row below is small.

The counts sum to 472. Shares are rounded to one decimal place and therefore do not sum to exactly 100.

One quarter of the sample is about getting product data across correctly. Products and variants took 121 of 472 threads (25.6 per cent), which is more than pricing, tax and inventory put together. For a mid-market manufacturer or distributor, that is the finding worth carrying into a platform decision: the connector work that eats the schedule is usually catalogue shape, not the order flow everyone worries about first.

The errors row looks small at 26 threads, and it is small only because of the precedence rule. Counted as a cross-cutting flag instead, 88 of the 472 titles (18.6 per cent) name an error, a failure, a defect, or something not working. So nearly one title in five is announcing that something broke, spread across the subject categories rather than concentrated in one.

Tax at 15 threads and inventory and warehouses at 18 are the smallest categories. We are not going to tell you tax and inventory are therefore easy on Acumatica. A public forum records the questions people choose to ask in public, and tax questions in particular tend to go to a VAR or to Avalara support instead. A low count here means low forum volume. It does not mean low incidence.

These are counts of thread titles matching a pattern, inside the same 472-thread sample read on 20 August 2026. They overlap with the categories above and with each other, so they do not sum to 472. Each row links to one thread we opened and read in full.

The matrix itemA matrix item is Acumatica’s way of describing a product that comes in variations, such as a shirt in four sizes and three colours. Acumatica calls the parent a template item and generates a stock item for each combination. group is the largest and the most concrete. Acumatica lets a template item carry more attributes than Shopify accepts as product options, and the connector reports that as a sync error rather than degrading gracefully.

Inside the same 472 titles read on 20 August 2026, Shopify is named in 194 (41.1 per cent), BigCommerce in 85 (18.0 per cent), and 12 titles name both. Amazon appears in eight. The remaining 205 titles (43.4 per cent) name no storefront at all, because they are about Acumatica screens, entity mapping or the Commerce Connector in general.

Do not read 194 against 85 as evidence that the Acumatica Shopify connector is worse than the BigCommerce one. We do not know how many Acumatica tenants run each connector, so we cannot turn thread counts into a rate. A connector with twice the installed base will produce roughly twice the forum traffic at identical quality. Acumatica does not publish the split, and we did not find a credible public source for it, so this comparison stays a raw count and nothing more.

One thing the split does support: if you are choosing between the two on Acumatica, the Shopify forum trail is deeper, which means more of your specific questions already have a public answer. That is a real advantage during an implementation and it has nothing to do with which connector is better built.

Read these before you cite the numbers above. Every one of them limits what the sample supports, and none of them is fixable by reading more index pages.

The line we will not cross: we are not going to write "X per cent of Acumatica users experience pricing problems". The sample is threads, the population is unknown, and no amount of arithmetic bridges that gap. Anyone who publishes that sentence from data like this has made it up.

Reproducibility is the whole point of this page. Follow these steps and you should land within a few threads of our numbers, allowing for topics posted since 20 August 2026. If you land somewhere very different, we want to know, because that means one of us has a bug.