DiscoveryOS Integration Requirements and Checklist | PA
Prepare a DiscoveryOS integration with catalogue mapping, events, storefront responsibilities, validation, rollback and operational ownership.
LEARN
Working with Particular Audience - FAQs
Published 8th Sep 2022 by Beth Smith
Make your website intuitive
Discover the Particular Audience platform.
Learn more
What is the integration process?
The integration process includes everything from signing the Statement of Work to getting our platform live on your website. Our sales team, account managers and engineers will work tirelessly to ensure that our platform is live as soon as possible.
How can I speed up the integration process?
Two simple ways to speed up the process are:
1\. Answer any questions from your sales manager as quickly as possible.
2\. Working with your sales manager to define what elements of the widget build are necessary for first release.
Certain elements of a widget build are more time-consuming than others. At the start of the integration process your sales manager will discuss which elements are likely to take longer to build, and which elements could be worked on post live date.
Why do you need a product feed?
We can secure a lot of customer behavioral data with our tracking tag, but the most accurate way of obtaining product data (e.g. availability, price, sale price, image url) is to use your product feed.
Using a feed means that product data will always be up to date with the latest information.
Why are test purchases needed?
We place test purchases to ensure that we are tracking customers through the checkout process correctly.
Why do you need a dispatch address and coupon code?
Typically we use your warehouse’s address as a delivery address. In the event of a test purchase not being successfully canceled, the order cannot leave the site if the warehouse address is the delivery address.
We require coupon codes to ensure we can correctly track a checkout even when customers add a coupon code.
What will you use historic sales data for?
When we go live, we will have comparatively little data on sales. Historic sales data is used to enrich our dataset and generate more relevant recommendations faster.
This will allow us to drive AOV more successfully as soon as possible.
Categories
[Learn\\ \\ \\ \\ \\
Our Products
Retail Media
Automating supplier sponsorship, retailers can use onsite real estate to generate incremental high-margin revenue and give trusted brand partners direct access to high intent consumers.
Learn more
Recommendations
Personalize every single product list for each unique customer using basket and comparison behavior, product images and item metadata.
Learn more
Price Beat
2x conversions with real time, MAP compliant competitor price beat technology that guarantees you always have the best price.
Learn more
Bundles
Automate product bundles using advanced AI. Drive up to 60% increase in Units Per Transaction, performance pricing available.
Learn more
Recent Posts
\\ \\
Press Particular Audience raises $10m to break down data silos in eCommerce and help retailers take on Amazon
\\ \\
Press TechCrunch Features Particular Audience Series A Funding
\\ \\
Press CEO and Founder James Taylor Shares the Benefit of Using Product Data NOT Customer Data to Engage Customers on Add to Cart
Plan your DiscoveryOS integration
Start with the systems and shopper journeys in scope. Agree how catalogue data, requests and events will move between them, who owns each boundary and how correct behaviour will be demonstrated before launch.
## Map the integration boundary Identify the catalogue source, commerce platform, storefront, event pipeline and any retained search or advertising components. Assign a technical owner to each system and document where identifiers are created and transformed.
## Prepare representative data Include products with variants, unavailable items, missing fields and market-specific conditions. Define which fields are authoritative and how deletions and updates are represented. Validate update semantics against the current API documentation rather than assuming that an omitted field is unchanged.
## Agree the event contract Define the event name, identifier, timestamp and relationship to a displayed product or offer. Specify deduplication and failure handling. Align the reporting definition with the event contract before using the data for an experiment or commercial reconciliation.
## Validate the storefront Test relevance, eligibility, labels, accessibility and the fallback when a request fails. Compare the displayed product, selected variant and cart outcome. Keep the old implementation available until the acceptance criteria and rollback procedure have been demonstrated.
## Handover operations Record monitoring, escalation, maintenance and release responsibilities. Set the launch sequence from observed readiness and dependencies; a generic timeline cannot establish readiness for a specific deployment.