Spatialedge — Opportunity Analysis
Last updated: March 27, 2026 (first call with Retief — Alex, Mike, Retief, Juan)
Sources: Call 1 (Mar 4, Mike + Juan) · Call 2 (Mar 6, Alex + Mike + Juan) · Call 3 (Mar 27, Alex + Mike + Retief + Juan) · Follow-up texts
Purpose: Evaluate whether Goose Group should pursue a commercial partnership with Spatialedge
Who they are
Spatialedge — South African data engineering and AI company. 10 years in operation. 170 employees — roughly 25% data engineers, 50% data engineering + ML engineering combined, remainder software engineers for platform development. 23 PhDs in data science, operations research, and mechanical engineering. Based in Stellenbosch.
Leadership: Retief Gerber (CEO), Jacques (CTO). Two capable board members. One investor who provided $5M specifically for international expansion.
Profitable and self-funded. No investor pressure. They've saturated the South African market — that's the driver for going international. Maintained 40% year-on-year growth while transitioning from consulting to product-led over the last 18 months.
They've built a talent pipeline that shouldn't exist. Engineers with 6-9 years of experience stay despite being offered double or triple pay elsewhere. The company has become the place to work in Stellenbosch — people graduate university and just don't leave.
Major clients in South Africa: Vodacom (10 years), Cell C, Woolworths, Take-a-lot, FNB, Lasaka, PepCor, Massmart. They've also landed two global brands operating in every single country in the world.
What Retief said (Mar 27 call)
This was the first direct conversation with Retief. Key things he said explicitly:
On the gap: "We are good at going and finding the technical problem and the technical solution. We suck at actually selling it to the client." — This was unprompted and direct. He knows the sales function needs sophistication to match the delivery capacity.
On existing clients: Realistic that there's limited headroom with current accounts. Vodacom won't accept outside strategic advisory — "they're never going to listen to you." Other clients are either already well-served or difficult to work with. "I don't want to create unrealistic expectation. I think looking at doing it together at new clients is obviously where the biggest prospect lies."
On what excites him: The platform partner path — Anthropic, AWS, Databricks. "This is also something that really excites me and also is the most obvious thing I think we need to take forward."
On Spatialedge's edge: "If there's one thing that Spatialedge is probably a world leader on, it's getting these non-deterministic solutions to work in an enterprise environment." Not the technology — the hard part is making it reliable, trustworthy, and actually used. They've shipped agentic use cases that give answers people can trust.
On the combined story: "If they work with us and give something to us, it's going to be number one not die as a POC. It's going to be something that gets used and the consumption is going to be driven up like crazy."
On ambition overlap: Spatialedge wants to focus on platform and delivery. GG focuses on commercial strategy and account growth. "I don't think there's too much in terms of our ambition that we overlap." — Clean separation of roles.
Product portfolio
DataLoom — Their agent product. An orchestrator with sub-agents, where each product is an optional app. This is the platform architecture.
DataCyst — Managed data pipeline product. Solves "95% of data problems really quickly." Becoming agentic — goal is to set up data pipelines from a CEO/CFO conversation in 30 minutes. Subscription revenue, tiered by pipeline count ($8k → $12.5k → $25k/month).
ML model lifecycle management — Build and deploy models to production in 30 minutes. Maintains them over time. Subscription service so the client doesn't need in-house ML ops.
Decision application stack — Rapid deployment environment for enterprise apps. Custom front-ends, decision workflows. Goes beyond dashboards.
Enterprise ChatGPT on private data — Tip-of-spear product for conversations. Easy to explain, gets people in the door.
Overall: 40 case studies, 3 generic infrastructure products, 12-14 client-facing products. A senior retail guy wants to spin out "Retail Edge" as a separate business.
Discovery team — Dedicated team whose job is to chase down any hint of a business case, scope it, define it, motivate it, and hand it to sales. They do free discovery → scoped proposal → implementation. Massmart (Walmart Africa) is currently paying for a discovery phase to package six different projects.
Pricing (confirmed by Retief)
Revenue stream Pricing Notes ────────────────────────────────────────────────────────────────────── Pod (3-4 people) $20-25k/week ~$100k/month all-in Platform fee $10-20k/month Tiered by pipeline count Maintenance/ops ~$20k/month "Keep the lights green" Typical engagement 6-12 weeks build Then maintenance phase
Three fee streams per engagement: consulting capacity, platform subscription, and managed operations. Retief acknowledges they can charge "a whole lot more" but are trying to get into the market.
US client feedback: the "keep the lights green" product sold before the project started — customer said equivalent headcount (3 people at $120-150k/year each) had been cut and they didn't think they could do it. Bought the product immediately.
Client updates (Mar 27)
Vodacom — 100M rand deal
10-year relationship. New use case every 3 days on average. Just signed a 100 million rand deal to transition Vodacom's entire stack onto Spatialedge's product platform. Started as staff augmentation, now fully product-led.
US client (transportation)
14 use cases identified from recent US visit. 6-month pipeline. Four projects running concurrently. On their sixth or seventh project. Insurance claims processing is a current workstream — they spend more money processing claims than the claims themselves cost. CSMs flag opportunities, discovery team scopes them.
European client
$300 million/year problem. 9-month engagement, constantly iterating and adding features.
Massmart (Walmart Africa)
Currently paying for discovery phase. Six problems identified by Spatialedge. Rough pricing given on all six, customer said yes to all — now getting into detail.
US deal: Revelist (San Diego)
Customer: Revelist — Rob's company. Rob is the ex-CEO of Juan's former company. Now runs a portfolio of 5 companies (golf simulators/fitting, outdoor/snowboard gear) acquired by Strategic Value Partners (private equity, $20B portfolio).
Started at $20k/week, now doubling. Rob's reaction when he heard the original price: "Why is it so cheap? Is it Mickey Mouse?" — he expected to pay more. Customer is comparing to:
Spatialedge $20k/week → doubling (12 specialists touching the project)
Deloitte comparison $60k/week (less complexity, less delivery)
McKinsey $3M deck, 12 guys on site ("dog shit" output per Rob)
Rob willing to cancel other projects and furlough engineers to prioritize Spatialedge work. Also doing: NetSuite data consolidation, sales data integration, record matching across 5 companies under the SVP umbrella.
The PE firm (SVP) has already come sniffing — "How are you guys doing this? We want to meet these guys." Direct path into a $20B portfolio of similar companies.
Europe deal: DHL (confidential)
Do not share this externally — confidential.
Free POC through CFO of DHL Spain — Juan's personal contact. The problem: revenue forecasting, currently done by one guy with a spreadsheet containing 15 pivot tables.
Could roll out to every country in Europe — "one month project, two week installation" per country. The CFO said "I'll take you to every country in Europe."
DHL has McKinsey permanently with hundreds of people on site — but this problem is "too small" for them. That's the wedge.
Priced at local approval limit to avoid going to the mothership for sign-off. Smart entry strategy.
Cloud Factory — updated picture
Cloud Factory (cloudfactorygroup.com) is Spatialedge's current US partner. Former annotation business trying to pivot into AI services. They're reselling Spatialedge without a product of their own.
What's changed:
Hired a CIO from the transport company (the one customer who bought 7 projects). Want to do whale hunting above a certain revenue level and want exclusivity. M&A talks failed — "books were bad." Cloud Factory won't acquire Spatialedge, and Spatialedge won't acquire Cloud Factory.
Also hired 5 senior sales people from Cloud 3 (acquired by Cognizant — ~1,000 engineers doing Azure migrations and AI services). Enterprise-level hunters.
Juan's position: Let Cloud Factory whale hunt above a threshold revenue level. Spatialedge retains the ability to sell below that threshold through other channels (Juan, GG, others). This is the compromise being negotiated.
Netherlands partner
A consulting group that installs stuff in companies. They wanted Benelux exclusivity.
Juan's position: "Anyone can have exclusivity on stuff they found first." Meaning: if they bring the deal, they own it. But no blanket territory lock-out.
Juan — clearer picture
Juan spent 12 months doing product strategy and go-to-market work for Spatialedge. Most of it was ignored — they kept trying to enable the Cloud Factory partnership instead. The turn happened recently: Spatialedge acknowledged Cloud Factory is broken and told Juan to do direct sales.
Juan's self-assessment (his words):
Good at: Gorilla sales, product sales, landing deals, customer success, account management. He's the tip-of-spear — gets the first deal in and runs it.
Not good at: Consultative sales, land-and-expand, pricing strategy, choosing between opportunities. He knows he's not the one to build and run a multi-channel sales operation.
What he wants: To work with/for GG, not build a team himself. He wants the structure and commercial horsepower that GG brings. He's not gatekeeping — "anyone can have exclusivity on stuff they found first."
Credibility with Spatialedge: After the Mar 4 call with Mike, Juan relayed the conversation to Spatialedge leadership. Their response: "Everything you said was true." Juan has built trust through 12 months of honest assessment, even when they didn't listen.
The partnership model (sharpened after Mar 27)
The Mar 27 call clarified the direction. The earlier framing — account expansion on existing clients, idea factory, agent agreement — has been narrowed. The path forward is new clients together, led by platform partnerships.
1. Platform partner path (primary)
Anthropic, AWS, Databricks all have account teams sitting with customers who want to do something real with AI. Their default referral is Accenture or McKinsey — 6 months of scoping, a team of 30, a deck. Those account teams are "fucking scrambling" (Mike's words from AWS experience, confirmed by Retief). They need someone who shows up with a use case, understands the customer's problem, and builds a working system in weeks. GG provides the commercial strategy and account planning. Spatialedge provides the engineering and platform. Together: pick a target customer → write the account plan → bring it to the platform partner → execute.
2. Combined story for Anthropic Europe
Alex has a contact at Anthropic's European operations. They're hiring their first people in Europe. Their implementation problem: more clients than they can handle, no confidence in McKinsey/Deloitte partners, not interested in POCs. The combined pitch: GG + Spatialedge = things don't die as POCs, they ship to production, consumption goes up, and the use case machine compounds.
3. Account planning together
Exchange existing account plans under NDA. Build on what Spatialedge already has — their case studies, capabilities, and client knowledge. Write joint account plans for European target customers. The muscle to build is: identify a target, understand the company, write the plan, match it to an existing platform partner account plan, approach together.
4. GG fulfillment channel
GG has its own customers who need operational systems built. Spatialedge engineers join those build teams directly. GG provides product leadership and customer relationships, Spatialedge provides the engineering. Real work, real revenue, and a way to see the collaboration in practice.
Revenue potential (updated)
Scenario Current pricing Target pricing ───────────────────────────────────────────────────────── Single project (3mo) $240k $750k+ GG commission (12%) $29k $90k 2 projects/year $58k $180k 4 projects/year $116k $360k
Pricing intel from Rob (Revelist): customer thought $20k/week was suspiciously cheap. Deloitte charges $60k/week for less. McKinsey charged $3M for "dog shit." The pricing gap is real and the market will bear significantly more.
The European recurring product (DHL-style revenue forecasting) could be even bigger — annuity commission on multi-country rollouts. One month project, two week installation per country.
What we still don't know
What does Jacques think? He wasn't on the Mar 27 call. Retief is aligned — Jacques needs to be brought in.
What do their account plans look like? Retief wants to exchange under NDA. We haven't seen their actual case studies, solution details, or account tactics yet.
What does the commercial arrangement look like? Commission, retainer, or both? Territory? Exclusivity terms? The Cloud Factory threshold? None of this was discussed on the call.
Can we verify delivery quality firsthand? Retief offered to demo products. We should see it.
Who are the first target customers in Europe? Mid-market ($100M-1B revenue), consumer-facing, not well-served by current AI implementation options. Retief mentioned retail, telco, financial services, healthcare, transportation. Need to pick specific companies and write account plans.
Next steps (agreed Mar 27)
1. Sign NDA and exchange account plans. Retief asked for this explicitly. We share ours, they share theirs. Build on what exists.
2. Build the combined story. A pitch document for platform partners (Anthropic, AWS, Databricks): here's who we are, here's what we do, here's a target customer we've already scoped. Retief: "We need to put this combined story to pitch to someone."
3. Product demo session. Retief offered to demo their platform stack — data pipeline system, decision apps, ML lifecycle. We should see it.
4. Identify European target customers. Write 2-3 joint account plans. Match to platform partner account teams. Approach together.
5. Bring Jacques in. Share the one-pager and schedule a follow-up that includes Jacques (CTO).
Note: Mike has also prepared a separate opportunity framing — worth reviewing alongside this analysis. A new one-pager for Retief and Jacques summarizes the platform partner direction.
Call 1 transcript (Mar 4, Mike + Juan) →