Inspire - AI Personalization Engine

What it is & the business problem it solves

Inspire - AI Personalization Engine (formerly known as Inspire by Building Blocks) is http://CM.com 's set of AI modules that turn your customer and product data into personal marketing: product recommendations, behaviour-based triggers (such as abandoned-cart reminders), customer profiles and predictions, and measurement of what your campaigns earn. On the rest of this page and in the module pages we simply call it Inspire. All modules work on both CDP and CXP.

Most retailers have plenty of data (website visits, carts, orders, email clicks) but struggle to use it consistently for every customer. Inspire reads that data from the http://CM.com Customer Data Platform (CDP, where customer profiles and their events are stored), calculates results on a schedule, and writes them back to the places your marketing tools already use: customer profile properties in CDP or in CXP (CM's customer experience platform), the Product Feed, and the Recommendations API. Marketers then use those results in email campaigns, workflows and segments, and Web Personalization.

How it works (

Collect. Your website tracking (CM's Webtracking) and your order data send standard events to CDP: product views, add-to-cart, remove-from-cart, begin-checkout and purchase lines. Your product catalogue lives in the Product Feed (also called the item feed).
Prepare. The Data module, the foundation every other module needs, loads these events and the product feed into Inspire every 30 minutes by default, keeping up to two years of history. See the Data page.
Calculate. Each module you use runs on its own schedule (a cron: a compact time pattern such as 0 4 * * *, meaning every day at 04:00) or after the module it depends on has finished. Most modules are not scheduled by default and only start running once a schedule is set; see Personalization Modules. For example, the Base Item Item Recommender learns which products belong together; Event Based Retargeting then uses that to recommend alternatives to an abandoned product.
Deliver. Results are written to:
- Customer profile properties in CDP or CXP, for example a customer's recommended products, RFM segment or abandoned items. Use them in segments, workflows, MMC email campaigns and Web Personalization merge tags.
- Product Feed attributes: per-product results such as "goes well with" lists or popularity counts, for product pages and carousels.
- Recommendations API: serves personal and contextual recommendation lists to Web Personalization when the page loads, with a fallback list for customers without personal recommendations.
- Analytics screens: attribution results and A/B test results.
Manage in the cockpit. The cockpit is the front-end tooling of Inspire. In it you manage module instances, adjust the business rules behind your predictions, and follow schedules, statistics and runs. How the cockpit works (expert settings, preview and publish) is explained in Personalization Modules.

Example use cases

Customer journeys that combine several modules:

Journey	Scenario	Channel	Required products / modules
Win back abandoned carts	A fashion retailer emails shoppers who left items in their cart, adding alternatives for each abandoned item.	MMC email campaign via CDP workflow	CDP or CXP, Webtracking, Product Feed, Data, Behavioral Trigger (cart items only) or Event Based Retargeting + Base Item Item Recommender (with recommendations)
Personal homepage	A DIY retailer shows each logged-in or recognised visitor a carousel of products picked for them; new visitors see popular products.	Web Personalization (Recommendations API)	CDP or CXP, Webtracking, Product Feed, Data, Base User Item Recommender, Personalized Product Recommendation, Web Personalization
Segment-based newsletters	A pet-food webshop splits its newsletter audience by RFM segment ("Champion", "At Risk") and by favourite brand, and reminds customers whose food is due to run out.	MMC email campaign	CDP or CXP, Data, RFM Scorer, Attribute Tagger, Recurring Purchase, MMC Email Campaigns
Prove the value	The same retailer measures which emails and messages led to purchases and A/B-tests a recommendation carousel against a control.	Attribution results; A/B Testing app	Attribution, A/B Testing, MMC Email Campaigns / Campaigns, Web Personalization

Packaging tier

MMC packages

MMC is sold in four packages: Basic, Advanced, Pro and Enterprise. Inspire is available from Pro upward; it is not included in Basic or Advanced.

MMC package	Inspire
Basic	Not included
Advanced	Not included
Pro	Modules marked Pro in the table below
Enterprise	All modules: those marked Pro and those marked Enterprise

Inside Inspire every module has a tier: Pro modules are part of the MMC Pro package (and, as we understand it, also of Enterprise); Enterprise modules are part of the MMC Enterprise package only. The tier of each module is shown in the table below. The Recommender Engine, near-real-time variants and Next Purchase Moment are included in Enterprise and are not sold separately.

Enterprise customers also receive all Pro modules.

Module overview

Module (customer name)	What it does	Tier	Status
Data	Loads events and the product feed; foundation for all modules	Pro (included wherever Inspire is available)	Available, required
Behavioral Trigger	Finds abandoned carts, checkouts and browsing (including price drop and abandoned category) and sends the items to CDP or CXP. Also available near-real-time.	Pro (near-real-time variant: Enterprise)	Available
Event Based Retargeting	Adds product recommendations to abandoned cart, checkout and browse triggers and to after-purchase triggers. Also available near-real-time.	Enterprise (near-real-time variant: Enterprise)	Available
Personalized Product Recommendation	Personal product lists per customer, with business rules and fallback	Enterprise	Available
Contextual Product Recommendation	"Goes well with" or "alternatives to" lists per product, written to the Product Feed	Enterprise	Available
Frequently Bought Product	Products a customer buys again and again	Pro	Available
Recurring Purchase	Predicts when a customer will next restock a product or product group	Enterprise	Available
Recommender Engine	The transformer-powered recommender (since June 2026), including the transformer-based personalized recommendation	Enterprise	Generally available, not a separate product
Base User Item Recommender	Engine behind Personalized Product Recommendation	Enterprise	Available (base module)
Base Item Item Recommender	Engine behind Contextual Product Recommendation and Event Based Retargeting	Enterprise	Available (base module)
RFM Scorer	Recency, Frequency, Monetary scores and segments per customer	Pro	Available
Attribute Tagger	Each customer's preferred brands, categories or other product attributes	Pro	Available
Item Enricher	Adds 7- and 30-day interaction and purchase counts to products	Pro	Available
Persona Generation	AI-discovered customer personas and assignment of customers to them	Pro	Beta
Next Purchase Moment (shown as "Timing Propensity" in the cockpit)	Probability that a customer buys within 7 to 365 days, plus a churn flag	Enterprise	Available
Attribution	Credits purchases to email and Campaigns messages (last touch + assisted)	Pro	Available
A/B Testing	Evaluates Web Personalization experiments	Pro	Available

Near-real-time. With Enterprise you can choose near-real-time instead of batch for Behavioral Trigger and Event Based Retargeting: the module then runs after every incremental data load of the Data module (roughly every 15 to 30 minutes, illustrative: depends on the configured schedule). Incremental loading must be switched on for the Data module (ask your CM contact).

Recommender Engine. The recommender has been powered by a transformer model since June 2026. It is generally available and included in Enterprise.