Dynamics 365 Commerce

We would like to request the availability of an AI-powered product recommendation feature in Dynamics 365 Commerce POS. This functionality would suggest additional or complementary products based on the items currently in the customer's basket during checkout.

Such a feature would support upselling strategies, increase the average transaction value, and enhance the in-store shopping experience. As there is currently no available solution offering this capability, we believe this is a valuable opportunity to bring intelligent product suggestions to POS and help retailers boost sales performance.

Background & Business Need

In retail environments, especially at the Point of Sale (POS), predictable credit memo number sequences can pose a significant fraud risk. As highlighted by a customer in a recent support interaction, if credit memo numbers follow a sequential pattern, a cashier could potentially guess and manually enter the next number in the sequence. This opens the door to unauthorized or fraudulent returns.

Proposed Enhancement

To mitigate this risk, we propose enhancing the credit memo number generation logic by introducing a randomized component. 

Benefits

• Reduces the risk of fraudulent returns by making memo numbers non-predictable.
• Enhances security and auditability at the POS.

Customer Impact

This request stems from a real-world concern raised by a customer actively using D365 Retail and Commerce. Addressing this would not only improve security but also demonstrate responsiveness to customer feedback and operational realities.

avoid using Person-type Customer accounts. We’re requesting your input on the feasibility of an alternate solution that leverages GAB contacts.

We proposed an approach where each user accessing the Online Store would be associated with a Customer account (Type = Person). These Person-type accounts are linked to their respective Invoice Accounts (Type = Organization) via the Customer hierarchy. Delivery address, lead time, and auto-charge logic are driven by values defined on the Organization-type Customer account.

We do not want to create separate Person-type Customer accounts, as they believe this would:

• Introduce data clutter and confusion in their CRM
• Conflict with their existing customer/contact structure

They are proposing an alternative:

• Use contacts from the Global Address Book (GAB) linked to Organization-type customers
• Map these contacts to Commerce identities (based on email)
• Let users select a delivery address during checkout, and base lead time/charge logic on that selection rather than the user’s customer record

Idea:

1. Commerce Identity Mapping
2. Can GAB contacts be authenticated in D365 Commerce as B2B users without being Person-type Customer accounts?
3. Lead Time & Pricing Logic
4. Is it technically feasible to:

• Use delivery address selection at checkout (regardless of user record)
• Drive lead time and charges logic based on the selected address, not the logged-in entity

1. Commerce Limitations
2. What, if any, limitations would we face in using contacts instead of customer accounts for:

• Storefront login and access
• Order history visibility
• Address validation and delivery charge logic

The e-commerce site is currently experiencing issues with Google Tags and Facebook Pixels not functioning as expected through the out-of-the-box (OOB) implementation. Despite the correct tag placement with Google, the tracking mechanisms are not triggering or capturing data as intended.

Ideally, the integration of these tracking features should be natively supported within the e-commerce platform, rather than relying on custom development. Enabling seamless compatibility with tools like Google Tag Manager and Facebook Pixel would enhance marketing analytics and reduce dependency on bespoke solutions.

Purge Commerce transactions feature does include the RetailEODStatementEventLog and RetailEODStatementControllerLog tables, but these tables grow so rapidity with each Commerce Transactional statement and do not get removed after the statement is posted, there are customers with over 3 billion records in the RetailEODStatementEventLog alone.

The Purge Commerce transaction feature does not allow you to purge sales transactions newer than the previous calendar year. I have seen over 24 million records in the RetailEODStatementEventLog for 6 months of data. If you have 1 year of data is the RetailEOD log files, the files are still very large.

There needs to be a purge just for the log tables alone. This way customers who do not want to purge Commerce sales transactions, can still purge the RetailEOD log tables if they want.

Currently, when a cashier scans a coupon or enters a coupon code at the POS, and the transaction does not meet the coupon’s requirements, no message or feedback is provided. Displaying an error or informational message explaining why the coupon cannot be applied would help reduce confusion and save time at checkout.

Back In the mid 2010s there was a console called the steam machine, the intention to be to bring console gaming directly to the living room, in a small form factor with a controller friendly OS. With computers now being offered at smaller and smaller form factors, this dream is far from dead, and it’s obvious that this space is still very much an underdeveloped segment of the market. I want to suggest to you to develop a windows based Xbox os, a true operating system that is controller friendly, and a potentially cheaper alternative to windows. This OS would meet all the needs of someone building a pc solely for gaming, offering general web browsing, a robust Xbox based UI to allow for control via a controller, and being windows based ensures compatibility with all current games and applications. The intention is to bring the ideals of console gaming into the hands of pc enthusiasts. Some people just want the opportunity to build with their hands, and put together their own hardware, so why shouldn’t we enhance their experience and offer an operating system alternative, at a more affordable price, specifically for their needs. Thank you for reading this far, I feel this is a smart move in the space, as while Linux based operating systems have purpose built versions for this, comparability can still be an issue in some cases and to have an official operating system in this capacity would be to quite literally own the market.

Currently, D365 provides the option to generate the 'Channel Price Report Job' print archive only. It would be very beneficial to have the capability to choose a print destination for this report. This would enable users to send the generated report via email as part of a batch process. This report contains key information, such as price changes, that must be reported to product managers and the administration team, who can be notified via email.

For Store Managers and Loss Prevention - it's important for a store to have a clear understanding of their entire cash picture at the close of day. Currently, as each shift closes throughout the day, a Z report is generated that shows the over/short for a particular drawer/shift. Therefore, store managers have to collect these Z reports and manually figure out the total over/short for the day across the shifts/cashiers/drawers. They also need to see a total cash and check collected for the day to help them determine their bank deposit amounts. I don'[t believe this information can be easily obtained today by the store manager today unless they hold onto the Z reports of each shift and manually add them together at the end of the day.

While D365 will allow users to see a total cash, check and total over/short on the ERP side (statement posting screens) it's important to note that most store managers do not use the backoffice screens nor have access to the statement posting form as these are finance related functions.

Loss prevention is typically also wanting to be notified by end of day how much total over/short the store had in terms of cash and may want store managers to inform them of this information as part of end of day process - therefore waiting for statement posting and going to finance for the data causes time delays that could impede/delay the investigation.

It does not appear that the POS easily provides the store manager a proper "end of day" cash, check and over/short report - this could essentially be a consolidated Z report that recaps all the shifts closed during the defined "day" and displays the overall/net over/short by payment type for the store for the day.

p.s. - please ensure Ashish sees this :)

hi, my name is ali and my idea is called clear view. its an prescription-adjustable monitor. 2.5 billion people need glasses but don’t wear them, straining their eyes daily on screens, phones, tablets, computers and monitors and many more. its a monitor that dynamically adjusts to a users prescription by using using AI and micro-optics, so they can watch movies, game, or work without glasses or contacts. Microsoft’s AI (like Co-pilot vision models) can accelerate the calibration, making this a natural fit for Surface, Xbox, or Azure partnerships.roof of Concept (Based on Existing Tech)

proof of concept :

1.) AI Calibration:

• how it works : User takes a quick eye test (via a Surface/Xbox camera + Azure AI vision analysis). Microsoft’s existing eye contact feature in Teams proves they can track eye focus.
• prototype : Modify an existing monitor with liquid crystal lenses (like DeepOptics’ 32" 4K prototype) + AI-driven software adjustment.

2.) Hardware Adjustments:

• example :Researchers at Stanford built a display that compensates for vision defects using light-field tech. Pair this with Microsoft’s AI for personalized profiles.

3.) Microsoft’s Advantage:

• AI: Use Azure Vision AI to map prescriptions from selfie tests (like EyeQuest).
• Gaming/Productivity: Integrate with Xbox (reduce eye fatigue for gamers) or Surface (enterprise/education markets).

i would love to be able to make money from ideas like this because ive got them all the time.