The eCommerce RFP Template That Actually Gets Useful Proposals

Migration RFPs in eCommerce are generally built to compare price quotes, which is the wrong approach. The job of an RFP is to find the vendor most likely to deliver a clean migration on a project where roughly half the requirements will change in the first 60 days.

A 2023 Forrester study commissioned by commercetools found that 77 percent of replatforming projects exceed their initial timelines. The RFP is supposed to prevent that. Most of the time, it makes it worse.

Here is why. The usual reaction is to make the RFP longer, and a long feature list just gets a wall of yes from every vendor. You cannot tell them apart, so you fall back on price. 

A short RFP with the right questions does the opposite. It shows you who can actually do the work.

If you sell both D2C and B2B, it is two RFPs in one document with clear variants for each. This guide gives you the structure I wish more brands sent us.

It also includes a free 29-page workbook  (PDF or editable Google Docs), an AI prompt you can use to draft it in 20 minutes, and a scoring rubric you can use internally.

Key takeaways (TL;DR):

• Forrester found 77% of replatforming projects exceed their initial timelines. Most RFPs still select on price while ignoring the variables that drive those overruns.
• Data migration and integrations cause two-thirds of timeline overruns. They deserve more space in the RFP than features.
• D2C and B2B require different sections. Send the same RFP to both, and you will get proposals that miss the actual work.
• The questions that predict project success probe how a vendor handles ambiguity.
• AI prompts can draft a usable first version of your RFP in under 20 minutes.

1. Why most eCommerce RFPs select the wrong vendor

Most RFPs benchmark the wrong variables. They ask for hourly rates, team size, technology certifications, and case studies in similar industries. None of those predict a clean migration.

A 2024 Info-Tech analysis of sourcing outcomes found that requirements clarity was the single largest predictor of post-award success, more than vendor capability or price.

From my personal experience, brands write RFPs that list features and ask vendors to confirm they can build them.

The proposals come back almost identical. The brand picks the cheapest one. Six months later, the project is not done and over budget.

Nobody asked how each vendor handles the things you did not know to put on the list.

A useful RFP does the opposite. It asks vendors how they think, how they price uncertainty, and how they behave when reality and the scope diverge. 

In the past 4-5 years, we've done over 100 migrations to Shopify. Some of those projects were also what we call in the industry rescue projects, where we took over someone's work and either cleaned up the code or started from scratch.

Here are some observations: 

The categories that cause Shopify migrations to go over budget and miss launch deadlines are not the ones most RFPs focus on. Data and integrations alone account for nearly two-thirds, based on our experience.

2. The 6 sections every eCommerce migration RFP needs

Every Shopify migration RFP, whether D2C or B2B, should contain six sections in this order:

1. Business context and migration drivers
2. Technical scope and required functionality
3. Data migration plan
4. Integrations inventory
5. Launch, QA, and risk plan
6. Commercial terms

If a section is missing, the proposals you get back will be either generic or wrong. If two sections collapse into one, the vendor has to guess. Vendors/agencies who guess price defensively, and you pay for it.

Whatever path you take, keep the order. Order signals priority, and you want the technical and data sections to dominate before you talk about money.

For a broader migration context, see the Shero guide on migrating to Shopify.

The downloadable Shopify Migration RFP Template (at the end of this post). The structure is identical for D2C and B2B brands. The content inside each section is what changes.

3. Section 1. Business context and migration drivers

Most brands skip this section or condense it into a single paragraph. That is a mistake.

The business context is what lets the vendor scope the actual project rather than the surface request. Write three to five paragraphs answering the following questions.

• What is the trailing twelve-month GMV?
• What is the growth rate you are projecting?
• What are you trying to fix or unlock by migrating?
• What is non-negotiable?
• What is the political reality on your buying team?

D2C variant

For D2C brands, add: average order value, repeat rate, top three traffic sources, mobile share of revenue, current conversion rate, and your CRO and email program maturity. If you run subscriptions, say so on page one.

B2B variant

For B2B brands, add: number of active business accounts, percentage of revenue from top 10 accounts, current ordering channels (rep, portal, EDI, punchout, phone), credit and net terms model, and whether you operate any hybrid D2C revenue alongside it. See the Shero Shopify B2B guide for the full scope of B2B features. If you serve public-sector or healthcare buyers, flag that early, because procurement constraints will shape the whole project.

Shopify migration timelines by source platform and business model. B2B migrations average 30 to 50 percent longer than D2C from the same source platform. Most RFP timelines undercount this.

4. Section 2. Technical scope and required functionality

This is where most RFPs go wrong by being too detailed. A feature list ten pages deep makes the proposals look identical because every vendor will write "yes" next to every line item. Instead, list the half dozen functions where you have an opinion about how they should work.

D2C variant

 Focus on theme architecture (custom build vs Shopify 2.0 template), checkout customization scope, ESP and CDP integration, loyalty, reviews, subscriptions, and the storefront performance budget.

Specify your target Lighthouse and Core Web Vitals numbers. They will rule out half the candidate themes immediately.

B2B variant

Focus company and contact data model, customer-specific catalogs, customer-specific pricing, net terms and credit holds, quote-to-order, draft orders for reps, punchout integration (cXML or OCI), EDI inbound, and approval workflows.

Punchout, CPQ, and net terms

If any of these are required, dedicate a subsection to them. Vendors who have not actually delivered cXML punchout will write "yes" anyway. Asking them to describe their last three implementations forces specifics.

Custom catalogs and pricing

Not every platform can handle everything, and ultimately you'll have to compromise.

Describe your current pricing logic before asking the vendor how they would model it. You will learn more from how they translate your current logic than from how they describe Shopify's capabilities (if you are considering Shopify).

5. Section 3. Data migration plan

This is the section that determines whether your project launches on time and on budget.

For example, a Shopify Plus enterprise migration can take 12 to 24 weeks. According to Shopify's migration guidance, data migration absorbs the largest share of delays.

The reason is rarely the migration tooling. It is that the source data is worse than the brand assumed.

In the RFP, ask vendors to describe their approach to four specific scenarios.

• Customers with duplicate emails.
• Orders with missing line items.
• Products with inconsistent variant structures.
• Historical order data that does not reconcile with your ERP.

Then ask them how they price each scenario. A vendor who quotes a fixed price for "data migration" without seeing your data is either inexperienced or hiding a change-order strategy.

On a recent Magento-to-Shopify migration project for a B2B brand on Magento 2, the customer data extract contained 41 percent email duplicates in the first sample. The client had assumed it was clean.

The data section of their RFP determined whether the vendor caught that in week 2 or week 14. A migration audit before you write the scope moves that discovery even earlier.

6. Section 4. Integrations inventory

List every integration. Every connector, every webhook, every nightly sync. The boring ones too.

D2C variant

Some common integrations are: ESP (Klaviyo, Attentive), CDP (Segment, mParticle), reviews (Yotpo, Okendo), subscriptions, loyalty, 3PL or warehouse, tax (Avalara, TaxJar), fraud, helpdesk, attribution.

B2B variant

For B2B-heavy clients, we see integrations such as ERP (NetSuite, SAP, Microsoft Dynamics), OMS, PIM (Akeneo, Salsify), CPQ, EDI translator, punchout middleware, customer portal authentication (SSO, SAML), and payment terms providers (Resolve, Slope, Balance, TreviPay).

Two columns. Required at launch versus phase 2. Vendors will price required and ignore phase 2. That is fine if your buying team agrees on what is required before issuing the RFP.

7. Section 5. Launch, QA, and risk plan

Most RFPs do not ask vendors what they will do if something goes wrong. Ask anyway.

Specify your launch model. Big bang cutover. Soft launch by traffic percentage. Geography by geography. Each implies a different risk profile and a different test plan.

Ask vendors to describe their rollback plan. Ask what triggers it.

A vendor with a real rollback plan has had one fail.  Ask about their post-launch support. A vendor without one has either never launched something critical or is willing to lie on an RFP.

Ask about their performance budget and how they enforce it. The Shopify themes that score 95 Lighthouse on a clean dev store can drop to 42 once your apps, custom blocks, and tracking pixels load. The Shero speed benchmarks study walks through the typical regression pattern.

8. Section 6. Commercial terms and what NOT to lock in

This section is short. That is the point.

Specify the structure. Fixed fee for discovery. Time and materials or a fixed fee for the build. Capped change order rates. Defined retainer after launch.

Do not lock in a total project price before the discovery phase. Any vendor who agrees to a fixed total without seeing your data is either lying or padding. Either way, you lose.

Specify your invoicing terms, your acceptance criteria for milestones, and what happens if milestones are not met.

For context on real Shopify Plus build pricing, see our Shopify pricing guide.

9. D2C vs B2B RFP: what changes in every section?

Same six sections. Different content. Different scoring weights. This table is the cheat sheet to keep both variants honest.

The weighted scoring rubric we use to evaluate vendor responses. D2C and B2B weights differ on integrations and CPQ/punchout but share the same total of 100.

10. The 8 questions that actually predict success

A Deloitte survey of global outsourcing has repeatedly found that implementation breakdowns, more than the initial vendor choice, drive most failed engagements.

The questions below predict implementation behavior. They reward vendors who have actually delivered projects like yours, not the ones with the most polished sales deck.

Predictive signal of each RFP question type, scored on our internal framework. The questions large agencies prefer (team size, client lists) sit at the bottom. The questions that scare junior teams sit at the top.

1. How do you price unknowns?
2. Walk us through a project that went badly. What did you learn?
3. How do you handle data that turns out worse than expected?
4. Who specifically will lead our project? Show us their last three.
5. Show us three references in our category with similar complexity.
6. What is the smallest project you would take? Why?
7. What is your most common change order pattern? Why does it happen?
8. What is one thing you would push back on in our RFP?

The AI prompt that drafts the RFP for you

Drop your brand context into the prompt below and run it in Claude (Sonnet 4.6 or Opus 4.8 handles the nuance best) or ChatGPT. Both produce a usable first draft in 5 to 20 minutes. Edit, then send.

The prompt running in Claude. Output a usable first draft of a Shopify migration RFP from a single message containing your brand context.

11. Red flags in vendor responses (and when RFPs are the wrong tool)

When proposals come back, here's what to look for and some red flags.

Identical case studies. If three different proposals describe similar past projects with the same vague metrics ("conversion up X percent, traffic up Y percent"), the vendors have not done your kind of project. They are mapping your RFP onto their generic deck.

Fixed total price before discovery. Already covered. Walk away.

A team size brag. If the vendor leads with team size, they are protecting against missed deadlines with bodies instead of careful scoping. That model breaks on complex B2B migrations.

No questions back. The best vendors come back with 10 to 20 clarifying questions in the first week.

The worst ones answer everything in the RFP exactly as written. The first group reads the RFP. The second group fills it out.

Why an RFP might not be the right approach

Sometimes an RFP is the wrong tool entirely. CIO.com argues that the traditional software RFP is broken for most modern buyers. They are right more often than buying teams want to admit.

If your project is under $80,000, the cost of running a proper RFP process exceeds the optionality it buys you. Pick two referenced vendors and run paid two-week discovery instead.

If you already trust an existing partner, do not run an RFP just because procurement asked. The hidden cost is the partner you lose during the process.

The RFP is most valuable on builds between $150,000 and $1 million where the field of qualified vendors is real and the price spread is wide. Below that range, references and a paid discovery are better options.

12. What to do after you receive proposals

The vendor that scored highest in the rubric is rarely the one that wins the project. The vendor that scored highest, interviewed best, and delivered the cleanest paid discovery is.

Run three steps after proposals arrive.

Step 1. Score each proposal against the rubric in Appendix A. Eliminate anything below 70.

Step 2. Interview the top 3 to 4. Give them the same case study problem from your real data. See what they ask, what they ignore, what they push back on.

Step 3. Pay the top 2 to run a 2 to 4-week discovery. Same scope. Compare what they find, what they cost it at, and how they communicate.

Then choose. The right answer at this stage is almost always obvious because you have replaced the abstract proposal with two real deliverables.

Conclusion

By Q3 2027, more than half of mid-market eCommerce migration RFPs will be AI-drafted on the buyer side. The polished, boilerplate-heavy RFP will get easier to produce. The differentiator will move further toward the paid discovery phase.

Adapt your selection process to shorter RFPs and longer discovery, and you will get better candidates and on-time, on-budget delivered projects. 

Frequently asked questions