# Who to hire for PredictSpring, now Salesforce Point of Sale PredictSpring became Salesforce Point of Sale, and a store rollout is part retail operations and part commerce integration. Hire for both halves, and make the partner prove it on a pilot store. **Published:** 2026-06-04 **Canonical:** https://synconai.com/hire-predictspring-salesforce-partner **Author:** SynconAI Architecture Team ## In short PredictSpring completed its move to Salesforce on 12 September 2024, and Salesforce's help centre now calls the product Salesforce Point of Sale, formerly Retail Cloud Modern POS and PredictSpring. Hire a partner who can run store rollouts as well as integrate with B2C Commerce, Order Management and Service Cloud, and who proves it with a pilot store. ## Sources - **Salesforce Signs Definitive Agreement to Acquire PredictSpring** (retrieved 2024-09-12) https://www.salesforce.com/news/stories/salesforce-signs-definitive-agreement-to-acquire-predictspring/ > Salesforce completed its acquisition of PredictSpring on September 12, 2024. - **Salesforce Signs Definitive Agreement to Acquire PredictSpring** (retrieved 2026-10-13) https://www.salesforce.com/news/stories/salesforce-signs-definitive-agreement-to-acquire-predictspring/ > PredictSpring’s Cloud POS solutions enable store associates to engage shoppers and complete transactions from anywhere in the store via a mobile device. ## What to take away - PredictSpring completed on 12 September 2024 and is now Salesforce Point of Sale, so search and contract under the current name. - The work has two halves: store operations such as devices, payments and training, and integration with B2C Commerce, Order Management and Service Cloud. - Salesforce lists no point-of-sale credential on its consultant, developer or architect pages, so judge the partner on store go-lives, plus B2C Commerce and integration certifications. - Scope a pilot store first, with a written plan for catalogue, inventory, returns and tender types before any wave rollout. ## The guide A retail point-of-sale project fails in a car park at seven in the morning, when a store manager cannot take a payment and the person who configured the system is in another time zone. That image is worth holding on to while you read proposals, because the partners who look identical on a slide differ most in how they handle the store. Hiring for PredictSpring is hiring for a rollout discipline first and a Salesforce skill set second. ## The product you are really buying The acquisition completed on 12 September 2024. Salesforce's announcement described PredictSpring as a retail point-of-sale vendor built natively in the cloud, and said its mobile approach meant that "PredictSpring’s Cloud POS solutions enable store associates to engage shoppers and complete transactions from anywhere in the store via a mobile device." It also listed Clienteling, Endless Aisle, Store Fulfillment and Store Operations alongside the till, and noted the company was already integrated with Commerce Cloud and Service Cloud. The full text is on the [Salesforce newsroom](https://www.salesforce.com/news/stories/salesforce-signs-definitive-agreement-to-acquire-predictspring/). Names have moved since. A Salesforce help article published on 3 April 2026 refers to "Salesforce Point of Sale (formerly known as Retail Cloud Modern POS and PredictSpring)", and the [Agentforce Commerce release](https://www.salesforce.com/news/stories/agentforce-commerce-announcement/) of 24 June 2026 presents point of sale as one part of a platform that also covers B2C Commerce, B2B Commerce and Order Management. That matters for hiring in two ways. A partner who still pitches "PredictSpring" with no mention of the current name may be working from old material. And the scope you write should name the current product and the editions and add-ons you have actually bought, because the commercial detail sits with your Salesforce account team, not with the partner. ## Two jobs in one proposal Read every proposal and ask which of two jobs it is pricing. The first job is the store. Devices, payment terminals, receipt printing, barcode scanning, network readiness, associate training, a cutover plan, store-by-store support in the first week, and a decision about what staff do when something stops working. The Agentforce Commerce announcement claims POS can launch "in as little as 18 weeks" and that store setup falls from a day and a half to under an hour. Those are Salesforce's figures, they depend on the retailer, and they are a useful prompt for asking a partner what their own store setup takes and how many stores a team can open in a week. The second job is the seam with the rest of Salesforce. The same announcement describes one catalogue shared between POS and the storefront, and says associates can change or cancel an order from anywhere in the store, whether it was placed digitally or in person. Making that real means agreeing which system owns product data, price, promotions, stock, customer identity and order status, and building the integrations that keep them in step. A partner that is strong on stores but thin on commerce data will leave you with a till that works and a customer record nobody trusts. Very few firms are excellent at both. Ask who is doing which, whether subcontractors are involved, and who is accountable when the two meet. The [integration guide](/integrating-predictspring-with-salesforce) covers the technical seam in more detail. ## Credentials and skills worth checking We read the Trailhead consultant, developer and architect credential lists on 13 October 2026 and found no certification that is specific to point of sale. That is itself useful: it means a partner cannot prove point-of-sale ability with an exam, and you must ask for delivery evidence. Credentials that still matter for the commerce half are the Salesforce Certified B2C Commerce Developer and the Salesforce Certified B2C Solution Architect, the Salesforce Certified Platform Integration Architect for the interfaces, and the Salesforce Certified Platform Data Architect if you expect large product and customer volumes. The lists are on Trailhead's [developer](https://trailhead.salesforce.com/en/credentials/developeroverview) and [architect](https://trailhead.salesforce.com/en/credentials/architectoverview) pages. Beyond exams, look for retail-specific skill: someone who understands tender types, split payments, gift cards, tax by location, returns without a receipt, end-of-day reconciliation and the way your finance team closes a trading day. Ask for the person's retail background, not only their Salesforce history. ## A first engagement: the pilot store Scope one store, or a small cluster in one region, before you scope a rollout. A pilot worth buying includes: - A written data design: product and price source, catalogue and stock flows, customer matching rules, and which system holds the order of record. - The tender and return rules for that store, including the awkward cases such as cross-channel returns and partial refunds. - A device and network plan, with the hardware list, the supplier and the delivery date. - Training for associates and a store manager runbook, tested on people who were not in the design workshops. - A cutover plan with a rollback to the previous till for the pilot store only. - Two trading weeks of hypercare, with agreed measures such as transaction time, failed payments and stock variance. The Agentforce Commerce announcement says iOS was live and Android and Windows were due in August 2026. Whatever your retailer's hardware, ask the partner to confirm in writing which devices are supported on the date you plan to go live, and treat that as a dependency on Salesforce, not on the partner. ## Where the data design usually goes wrong Most point-of-sale projects that disappoint were not undone by the till. They were undone by the data behind it. Three patterns recur, and a good partner will raise them before you do. The first is product data that was never clean. A web storefront can hide a messy catalogue behind a handful of well-photographed hero products. A store associate scanning a barcode cannot. Variants, sizes, bundles and discontinued lines all surface in the first week, and each one needs an owner who can fix it the same day. Ask the partner how they audit the catalogue before the pilot and how many records they would expect to correct. The second is customer identity. Clienteling only works if the person at the counter is the same person your web account and service history know. Matching on email alone will miss people who give a different address in store, and matching too loosely will merge strangers. Ask what rules the partner uses, who approves them, and how a wrong match is reversed. Salesforce's announcement positioned the product as complementing existing Customer 360 capabilities, which is precisely why the matching rules deserve a named owner on your side. The third is stock accuracy. Endless Aisle and store fulfilment promise that a customer can buy something that is not on the shelf, and that promise is only as good as the inventory number. If stock counts are weekly and sales are continuous, the first oversold item will cost more goodwill than the feature earned. A partner who is serious will ask how often counts are taken, which locations feed the number, and what happens when a store ships an item to a customer. Underneath all three sits a plain question of ownership. Write down, for each domain, which system is the master, which systems read it, and how often. Then ask each candidate to criticise your table. The firms that find real problems in it are the ones who have run a rollout before. ## Red flags specific to point-of-sale work Treating the till as a web order. A store transaction and a storefront order do not behave identically, and a partner who describes both as "orders" with no mapping is skipping the hard design work. No plan for the bad day. Ask what a store does when the network drops, when a payment terminal fails, and when stock on screen disagrees with the shelf. You are not looking for a perfect answer, you are looking for an answer that has been rehearsed. A big-bang rollout proposed with no pilot. Chains that open fifty stores in a weekend usually discover their catalogue problems fifty times. A fixed price that excludes hardware, store visits and training, so the headline looks lower than the real cost. No named person responsible for the peak trading freeze. If your busiest season falls inside the timeline, the plan must show where change stops. ## Ten questions to ask 1. How many point-of-sale stores have you taken live, how many were on Salesforce Point of Sale or PredictSpring, and which had B2C Commerce behind them? 2. How long did your last pilot store take from signed scope to first trading day? 3. Who owns product, price and stock data in your usual design, and why? 4. How do you handle returns that cross channels, such as buying online and returning in store? 5. What is your plan for a store with no network for an hour? 6. Which devices and payment providers have you integrated, and who certified them? 7. How do you decide what stays in the till and what flows to Service Cloud for customer service? 8. What do associates need to be taught, and how do you test that they learned it? 9. Which parts of the work are done by your own staff and which by subcontractors? 10. What would make you recommend pausing the rollout after the pilot? ## Evidence, references and commercial terms Ask to speak to a store operations manager at a reference retailer, not only the project sponsor. Their view of the first week matters more than the steering committee's. Ask for the go-live checklist from a past store, with the retailer's name removed, and for the measures they track in hypercare. A partner who has no hypercare numbers has probably not watched a store trade. On commercial structure, a fixed fee suits the pilot because the scope is bounded and the deliverables can be listed. Rollout waves suit a unit price per store with a defined definition of done, provided hardware, travel and training are either included or priced separately and visibly. Time and materials is sensible for integration discovery. The statement of work should name the data owner for each domain, the supported devices, the pilot exit criteria, the support hours in hypercare, and who pays when a store slips because of a dependency on Salesforce or on hardware delivery. Licence charges are with Salesforce, and the [licence guide](/predictspring-salesforce-licence-cost) and the [cost guide](/predictspring-salesforce-implementation-cost) show how to keep the two lines apart. The [migration guide](/should-i-migrate-from-predictspring) is worth reading if your question is whether to move at all. | Area | What to ask for | Warning sign | | --- | --- | --- | | Store rollout | Pilot plan with rollback and hypercare measures | Fifty stores in the first wave | | Data ownership | Named system of record per domain | Everything is "synced" | | Devices | Written list of supported hardware and dates | "Any tablet will do" | | People | Retail operations lead on the team | Salesforce specialists only | | Pricing | Hardware, travel and training shown separately | One lump sum | ## When a small project can skip a partner A single-location retailer with a simple catalogue, one payment provider and an administrator who is comfortable with Salesforce configuration may manage a pilot store with guidance from Salesforce's own implementation resources. The test is whether you can describe your data design and your bad-day procedure on a page. If the business has several stores, online sales on B2C Commerce, or loyalty and returns across channels, the integration work is the part most worth paying for, and that is where an experienced partner earns the fee. The [what happened guide](/what-happened-to-predictspring) records the dates, and the [implementation guide](/predictspring-commerce-cloud) goes further into the commerce side. ## Questions ### Is PredictSpring still called PredictSpring? Not in Salesforce's current documentation. A help article published on 3 April 2026 calls it Salesforce Point of Sale, formerly known as Retail Cloud Modern POS and PredictSpring. Write the current name into your brief, and ask the partner which editions and add-ons your contract covers, because that detail sits with your Salesforce account team. ### Which certifications show a partner can deliver Salesforce Point of Sale? None is specific to point of sale on Trailhead's consultant, developer or architect lists, read on 13 October 2026. Ask instead for B2C Commerce Developer, B2C Solution Architect and Platform Integration Architect credentials, then insist on store go-live evidence, with a reference who ran store operations. ### What should a pilot store include? A written data design, tender and returns rules, a device and network plan, associate training, a cutover with rollback for that store only, and two trading weeks of hypercare with agreed measures. It should finish with exit criteria that decide whether the wave rollout proceeds. ### How long does a Salesforce Point of Sale rollout take? Salesforce's June 2026 Agentforce Commerce announcement says POS can launch in as little as 18 weeks. That is a vendor figure and depends on the retailer, so ask each partner for the elapsed time of their own last pilot and what slowed it down. ### Can I hire one partner for the stores and another for the integration? Yes, and it is common. Name one party accountable for the seam between them, with a shared data-ownership document and a joint test plan. Without that, each firm's work can pass its own acceptance test while the customer record and stock figures still disagree between till and storefront. --- SynconAI is a Salesforce Select Partner and an OpenAI Select Partner, delivering architect-led Salesforce and AI programmes across the United States and Australia. Related service: Salesforce consultancy (https://synconai.com/salesforce-consultancy).