Most technology advice in the Kingdom is paid for twice: once on the advisor’s invoice, and again in the margin they earn on whatever they told you to buy. This post covers where that bias enters, the one question that exposes it, what an independent evaluation hands you, and how to run one on your next big purchase.
That second payment does not make reseller-consultants dishonest. It tilts their advice toward the platforms that pay them, and the tilt is invisible exactly when you most need a level scale: a seven-figure ERP, core system, or cloud contract on the table.
Where the bias gets in: four stages of a tilted evaluation
Every evaluation runs through four stages, and a vendor incentive can bend each one. It rarely looks like a lie; it looks like a sensible default.
flowchart LR A(["1. Requirements"]) --> B(["2. Shortlist"]) B --> C(["3. Scoring"]) C --> D(["4. Recommendation"]) A -.-> A2(["Written to match the partner product's feature list"]) B -.-> B2(["Built from the partner roster, not the market"]) C -.-> C2(["Weights favour what the partner does well"]) D -.-> D2(["Total cost stops at the licence; exit cost never appears"])
- Requirements. If the advisor who helps you write them resells one product, they arrive shaped like that product’s brochure, and read as perfectly reasonable.
- Shortlist. A partner roster becomes “the market.” The open-source option, the regional vendor, and “do nothing yet” quietly never appear.
- Scoring. Weighting is where a sober spreadsheet gets tilted. Give “vendor ecosystem” 20% and “cost to leave” 0% and the market leader wins every time. A fairer approach is in our software proposal scoring framework.
- Recommendation. Total cost stops at the licence and year-one services. Migration, integration, years two to five, and the cost of leaving are yours to discover.
The one question that sorts advisors from salespeople
Here it is as a picture, because it is this simple.
flowchart TD
A(["Your advisor recommends a platform"]) --> B{"Do they earn anything if you buy it?"}
B -->|No| C(["Real advisor. They only win when your decision is right."])
B -->|Yes| D(["Salesperson with a clipboard. The bias is baked in."])
D --> E(["Ask for the partner tier, the margin and any referral fee, in writing"])
Ask any advisor, in the first meeting: “Do you earn anything if we buy what you recommend?” If the answer is yes, you have a salesperson with a clipboard, not an advisor. That is no reason to show them the door; resellers know their product and implement it well. It is a reason to stop treating their comparison as one, and to ask, in writing, for their partner tier, margin, and any referral fee. In our experience the honest ones answer in a sentence.
A firm that resells nothing and takes no commissions only wins by getting your decision right. That changes which options make the shortlist, how honestly the trade-offs are presented, and whether anyone tells you the cheapest path is sometimes not to buy at all.
What a vendor-neutral evaluation actually hands you
Independence is a property of the process, not a claim on a website. A real evaluation produces five documents.
- A requirements document you own. Written from your workflows and constraints (data residency, integration with your existing ERP, Arabic support, audit needs) before any vendor is named. If a vendor helped write it, it is theirs.
- A shortlist with the rejected options attached. Including the regional vendor, the open-source route, “extend what we have,” and “wait a year.” A shortlist without rejected options is a sales pipeline.
- A weighted scoring matrix agreed before the demos. Weights fixed in advance, scores recorded per criterion, and the demo scripted around your hardest three scenarios, not the vendor’s best ten.
- A five-year total cost of ownership model. Licence, implementation, integration, migration, internal time, support, and the cost of leaving if it fails. The licence is the smallest number on that list; the rest is in the hidden costs of enterprise software.
- A recommendation memo we will defend in front of your board. The reasoning, the scoring, the runner-up and why it lost, and what would change our mind.
The base rate is poor. McKinsey and Oxford studied more than 5,400 IT projects and found that large IT projects (initial budget over $15M) run on average 45% over budget, 7% over time, and deliver 56% less value than predicted (McKinsey, 2012). Much of that gap opens the day the option is chosen, not the day the project slips. A tilted shortlist and a licence-only cost model are two of the cheapest ways to open it.
A worked example: the platform that cost twice the licence
Say a mid-sized distributor in Riyadh is choosing a new ERP; the numbers are made up to show the shape.
A reseller-consultant recommends the platform they partner on. Licence: SAR 600,000 over five years. Implementation by their team: SAR 700,000. Total on the slide: SAR 1.3 million, against two products that were never serious contenders.
An independent evaluation starts from what the distributor needs: three warehouses, bilingual invoicing, e-invoicing integration, and a data-residency requirement from its largest customer. The shortlist adds a regional vendor and extending its existing finance system. Scoring weights “cost to leave” and “integration with what we have” at 15% each. The five-year model then adds what the slide left out: data migration (SAR 250,000), warehouse-scanner and bank integration (SAR 300,000), two staff seconded for eight months (SAR 400,000), support uplift after year one (SAR 200,000), and an exit cost if it fails (SAR 350,000). The “SAR 1.3 million” platform is now a SAR 2.8 million decision. The regional vendor scores slightly lower on features, comes in at SAR 1.8 million all-in, and can be left without re-keying five years of data.
Neither answer is automatically right. But the owner now sees both numbers, and the second was never going to be on the reseller’s slide. The shape is typical: Nucleus Research has pegged ERP implementation at roughly 3 to 4 times the licence, so a licence-led comparison is not really a comparison.
Why this matters more for Etimad and regulated buyers
Government entities buying through Etimad, and anyone supervised by SAMA or subject to the NCA’s Essential Cybersecurity Controls, have a second problem: the decision has to survive audit and challenge. “The consultant recommended it” is not defensible when the consultant earns a margin on it. An independent evaluation gives you a paper trail that holds up: requirements, scoring, rejected options and why, and a cost model that includes the exit.
Two Saudi specifics change shortlists more often than owners expect. Personal data sits under the PDPL, in effect since 14 September 2023 with SDAIA as regulator, so where a platform stores and processes data is a hard criterion, not a preference; a global vendor without an in-Kingdom option can drop off the list before pricing is discussed (Morgan Lewis, 2024). A formal tender needs vendor-neutral requirements on paper too, or it attracts the best salespeople, not the best fit; see how to write a technical RFP that gets you the right vendor.
Where independence goes wrong
Independence is easy to claim and easy to lose. The traps we see most often:
- The “independent” firm with a referral agreement. No resale, no logo on the website, but a finder’s fee on closed deals. Ask about referral fees specifically.
- The consultant with a favourite stack. Paid by nobody, but has delivered the same platform six times and will recommend a seventh because it is what they know. Ask what they recommended on their last three evaluations; three identical answers is a pattern.
- The free assessment. Somebody funded it, usually the vendor whose product it says you need. Treat it as a proposal, not an evaluation.
- Letting the vendor write the requirements after all. You hire an independent advisor, then let the incumbent vendor’s architect “help with the technical detail.” The tilt walks back in by the side door.
- Analysis for its own sake. Independence is no excuse for a six-month study; a mid-sized purchase should close in weeks, with a decision at the end.
How to start: one decision, one paper trail
The pattern that works is boring and reliable: pick the purchase on your desk with the largest five-year cost, not the most urgent one. Write the requirements yourself, or with someone who earns nothing from the answer. Ask every firm the one question and record the answers. Fix the scoring weights before the first demo. Insist on a five-year cost model with an exit line. Then decide, file the paper trail, and check the model against reality at twelve months. If it held up, you have a template for every later purchase. If not, you have learned, cheaply, where your organisation’s tilt comes from.
Weighing a platform decision and not sure whose advice to trust? SDCG’s technology assessment and due diligence work is independent by design. No resale, no commissions, no vendor referral fees. We evaluate the market for you and hand you a recommendation you can actually defend. Book a free 30-minute review and talk through the decision you’re weighing.