Intelligent Sales Settings
Intelligent Sales Settings is the Salesforce Setup page that turns on and configures Intelligent Sales, the image-recognition retail-execution capability inside Consumer Goods Cloud.
Definition
Intelligent Sales Settings is the Salesforce Setup page that turns on and configures Intelligent Sales, the image-recognition retail-execution capability inside Consumer Goods Cloud. Field reps use the Intelligent Sales mobile app to photograph a store shelf, and Einstein Object Detection compares that photo against the planned shelf layout to score on-shelf availability, product facings, and share of shelf. The Settings page is where an admin enables the feature, points it at the right object detection model, and grants reps and managers access through permission sets.
It is not the same thing as the Einstein sales-AI features on core Sales Cloud, such as Einstein Activity Capture or Einstein Opportunity Scoring. Intelligent Sales belongs to the Consumer Goods Cloud product line and is aimed at consumer packaged goods companies whose reps visit physical stores. The setup ties together three moving parts: the licenses and permission sets that authorize use, the image model that reads shelf photos, and the store-visit data that field teams capture on the road.
How Intelligent Sales reads a store shelf
What Intelligent Sales actually does
Retail execution is the work of getting the right products onto the right shelves, with the right promotions, at the right time. Consumer goods companies send field reps into grocery stores, pharmacies, and convenience stores to check that this is happening. Intelligent Sales is the part of Consumer Goods Cloud that uses a phone camera and image recognition to do that check fast. A rep walks up to a shelf, opens the Intelligent Sales mobile app, and takes a photo that Salesforce calls a shelfie. Einstein Object Detection then looks at the photo and identifies which products are present, how many facings each one has, and which expected products are missing. The output lands back on the visit record so the rep can fix problems on the spot, such as restocking a gap or correcting a misplaced item. Without this, a rep would count products by hand and eyeball the layout, which is slow and inconsistent across a large team. The Settings page is the admin control that makes all of this available to the field. It is the difference between a shelf audit that takes ten minutes and one that takes thirty seconds.
Planogram versus realogram
Two terms sit at the center of Intelligent Sales, and they are easy to mix up. A planogram is the desired layout, a sketch or reference image from the sales manager showing how products should sit on the shelf. A realogram is the actual shelf as it exists in the store, captured in the rep photo. Salesforce stores both as records on the Image object, the planogram as the standard and the realogram as the shelfie taken during the visit. Einstein Object Detection compares one against the other. The model returns the detected product count, the number of facings per product, and the share of shelf, which is the portion of shelf space a brand occupies relative to the total. From those raw numbers the system derives compliance indicators such as facings met, share of shelf achieved, and voids, meaning products that should be present but are not. A manager uploads the planogram and ties it to specific in-store locations ahead of time. When the rep takes the shelfie on site, the comparison runs against that location, so the score reflects what that particular store agreed to stock.
Where the image model fits
The intelligence in Intelligent Sales comes from Einstein Object Detection, an image-recognition model trained to spot consumer products on a shelf. Salesforce trains the model on example images of the products a company sells, so the detector learns each pack, label, and variant. During a visit the rep photo is sent to the model, which returns bounding boxes and labels for the items it recognizes. The app then tallies counts and facings and calculates share of shelf from the positions. Because the model is product-specific, the quality of detection depends heavily on the training images supplied during onboarding. Poor or sparse training data produces missed or misidentified products, which is the most common reason early pilots disappoint. The Intelligent Sales Settings area is where the admin connects the org to the right model and confirms it is active. Image recognition runs through the Retail Execution APIs, so the feature also depends on the org being provisioned for Consumer Goods Cloud rather than plain Sales Cloud. This is a managed-package and industry-cloud capability, not a toggle that appears in every Salesforce edition.
Licenses and permission sets
Intelligent Sales is gated behind Consumer Goods Cloud licensing and a set of permission sets, not a single checkbox. The org needs the Consumer Goods retail execution permission set licenses, and individual users need the matching permission sets assigned before the mobile app and image recognition appear for them. Salesforce documents a dedicated step, Assign Permission Sets for Intelligent Sales, precisely because the feature stays invisible to a rep whose permissions are missing. The permission sets split roughly along role lines, with one grant for field reps who capture shelfies and complete visits, and broader access for managers who set planograms and read analytics. A frequent setup mistake is enabling the feature in the Settings page, then forgetting the per-user permission assignment, so admins see the feature live while reps see nothing. Permission sets also changed shape over releases, with one model before Summer 24 and a revised model after, so an org upgrading from an older Consumer Goods install should re-check its assignments rather than assume the old grants still map correctly to the current feature.
The store visit it plugs into
Intelligent Sales does not run on its own. It sits inside the store-visit workflow of the Consumer Goods Cloud mobile app, which is built to work offline because stores often have weak signal. A rep plans a route, drives to a store, and opens a visit that contains a checklist of tasks. One of those tasks is the planogram check, and that is where the shelfie and image recognition come in. The rep photographs the shelf, the model scores it, and the result feeds compliance metrics for that visit. Other visit tasks can include taking orders, checking promotions, recording out-of-stock items, and answering survey questions. The app keeps a store cockpit, a 360-degree view of the account that shows past visits, orders, and performance, so the rep arrives prepared. Background sync pushes the captured data back to Salesforce once a connection returns. Because the photo evidence and the scores attach to the visit, managers get an objective record of shelf conditions across many stores rather than relying on the rep memory. That audit trail is the real payoff of wiring image recognition into the visit instead of treating it as a separate tool.
KPIs and the analytics it powers
The scores Intelligent Sales captures are not just for the rep in the aisle. They roll up into retail store KPIs and analytics dashboards that managers use to steer the field team. From planogram check tasks Salesforce derives assessment indicators such as facings, share of shelf, and voids, then aggregates them into store compliance views. Tableau CRM dashboards for Consumer Goods Cloud, sometimes still labeled Einstein Analytics, surface insights on store compliance, last visit, product performance, and which stores need attention. Managers can compare reps against the team average, spot top and bottom performers, and see how compliance correlates with sales at a store. New-product tracking shows how recently launched items perform over their first months on shelf. This closes the loop from a single shelf photo to a territory-level decision about where to send reps next. The Settings page indirectly governs the quality of all of this, because if the image model is misconfigured or permissions block capture, the dashboards downstream are built on thin or skewed data. Clean capture at the shelf is what makes the analytics trustworthy.
Rolling it out without surprises
A sensible rollout treats Intelligent Sales as a field-operations change, not just a Setup toggle. Start by confirming the org is on Consumer Goods Cloud and that the retail execution licenses are provisioned, because the Settings page may expose options the org cannot actually use. Load good training images for the products in scope so Einstein Object Detection has enough to learn from, then test the detection against real shelves before any rep relies on it. Assign the Intelligent Sales permission sets to a small pilot group of reps and at least one manager, and have them run live store visits to judge accuracy. Watch for the gap between what the model reports and what the rep sees, and refine the training set where it misses. Confirm the planograms are tied to the correct store locations, since a comparison against the wrong layout produces misleading compliance. Only after the pilot reports reliable scores should the permission sets widen to the full field team. Sequencing this way avoids the classic failure where every rep gets the feature at once, detection is shaky, and trust in the tool collapses before it has a chance to prove value.
How to enable Intelligent Sales
Enabling Intelligent Sales means turning the feature on in Setup, connecting the image-recognition model, and granting reps and managers access. The org must already be provisioned for Consumer Goods Cloud retail execution before these steps work.
- Confirm Consumer Goods Cloud and licensing
Verify the org has Consumer Goods Cloud retail execution provisioned and that the matching permission set licenses are available. Intelligent Sales does not appear on plain Sales Cloud, so check this before touching any toggle.
- Enable Intelligent Sales in Setup
In Setup, open the Intelligent Sales settings and turn the feature on. This makes the image-recognition capability and the related mobile experience available to provision for users.
- Connect the object detection model
Point the feature at the Einstein Object Detection model trained on your product catalog. Confirm the model is active and has been trained on representative example images of the products reps will photograph.
- Assign permission sets
Assign the Intelligent Sales permission sets to field reps and to managers, following the Assign Permission Sets for Intelligent Sales guidance. Reps who capture shelfies and managers who read analytics get different grants.
- Tie planograms to store locations
Have managers upload planogram reference images and associate them with the correct in-store locations, so the comparison during a visit scores against the layout that store agreed to stock.
- Pilot, then expand
Run live store visits with a small pilot group, check detection accuracy against real shelves, refine the training images where the model misses, then widen the permission set assignment to the full team.
The industry-cloud base that Intelligent Sales runs on; required before the feature works.
The image-recognition model, trained on your product images, that reads shelfies and returns counts, facings, and share of shelf.
Per-user grants that make the mobile capture and analytics visible; separate grants for reps and managers.
The link between a reference layout and a physical store, which scopes each compliance comparison correctly.
- The Settings page can show the feature as on while reps see nothing, because the per-user permission sets were never assigned. Enabling and granting are two separate steps.
- Detection quality depends on the training images. Sparse or low-quality product images cause missed and misidentified items, which sinks early pilots.
- Intelligent Sales requires Consumer Goods Cloud. It is not the Einstein sales-AI bundle on core Sales Cloud and will not appear in a standard Sales Cloud org.
- Permission sets changed between the pre Summer 24 and current models, so an upgraded org should re-verify assignments rather than trust legacy grants.
Trust & references
Cross-checked against the following references.
- Enable Intelligent SalesSalesforce
- Assign Permission Sets for Intelligent SalesSalesforce
Straight from the source - Salesforce's reference material on Intelligent Sales Settings.
- Use Intelligent SalesSalesforce
- Introduction to the Intelligent Sales Mobile AppSalesforce
Hands-on resources to go deeper on Intelligent Sales Settings.
About the Author
Dipojjal Chakrabarti is a B2C Solution Architect with 29 Salesforce certifications and over 13 years in the Salesforce ecosystem. He runs salesforcedictionary.com to help admins, developers, architects, and cert/interview candidates sharpen their fundamentals. More about Dipojjal.
Test your knowledge
Q1. Why is understanding Intelligent Sales Settings important for Salesforce admins?
Q2. What is the primary benefit of Intelligent Sales Settings for Salesforce administrators?
Q3. In which area of Salesforce would you typically find Intelligent Sales Settings?
Discussion
Loading discussion…