Should I build our mobile app in-house or outsource to an agency?
Should you build a mobile app in-house or outsource? Validate demand first with a 6-week sprint before committing £100k. Performance-based agency contracts reduce risk.
Last updated:
Who This Is For
You're a B2B SaaS founder at £600k ARR facing real competitive pressure—your customers are explicitly asking for a mobile app and you're losing deals because of it. Your team is entirely backend and web focused with zero mobile experience, and your runway is tight enough that the 6-8 month in-house build timeline is genuinely risky, especially when an agency could deliver in 4 months for roughly what you'd spend hiring one mobile developer. This decision matters most right now because speed to market is directly tied to your ability to win deals and retain customers.
What the Board Says
"Reject both pure agency and pure in-house approaches. Instead: (1) Spend £15-25k on 6-week validation sprint (landing page, customer interviews, pre-sales) in Months 1-2; (2) If validation shows £20k+ revenue potential by Month 8, proceed with agency build (£80-100k, Months 3-6); (3) If validation is weak, pivot or shut down, preserving £125-135k runway. Include decision gate at Week 6: proceed to agency only if you secure 10+ paid pilot commitments or £15k+ in pre-sales."
Amara Osei "Neither pure approach is optimal."
"Neither pure approach is optimal."
"Neither pure approach is optimal. Recommend hybrid: (1) Hire 1 senior full-stack developer immediately (£5-6k/month, 2-month commitment = £10-12k); (2) Contract specialized agency for design/UX only (£15-25k, Month 1-2); (3) Developer builds core MVP with agency designs (Month 2-5, total £15-18k salary); (4) At Month 5, evaluate: if validation positive, hire second developer (£40k remaining runway supports 3-4 months dual team); if validation negative, clean exit at £40-55k total spend vs £100k agency sunk cost."
Henrik Sørensen "Pause for 48-hour discovery sprint before committing."
"Pause for 48-hour discovery sprint before committing."
"Pause for 48-hour discovery sprint before committing. Assign tasks: (1) Sales quantifies monthly deal loss rate and revenue impact; (2) Finance models 12-month costs for agency vs in-house vs hybrid; (3) Engineering evaluates 3 agencies with code quality requirements and penalty clauses; (4) Leadership defines success metrics for 12-month evaluation. THEN decide agency vs in-house with data, not urgency. If discovery shows >$150K/month revenue loss: Engage agency with structured contract (4-month delivery, knowledge transfer, code ownership, quality gates, 20% penalty for delays). Simultaneously hire 1-2 mobile engineers to shadow agency work. If <$150K/month: Build in-house with contract mobile consultant for first 3 months to accelerate team learning."
Maria Santos "Approve agency engagement with structured decision framework: (1) Quantify current revenue loss (spend 1 week with sales to calculate actual monthly deal loss rate and average deal size); (2) If monthly loss exceeds $150K, proceed with agency at $200K budget with penalty clauses (10% refund per week of delay beyond 4 months, 25% refund if crash rate >0."
"Approve agency engagement with structured decision framework: (1) Quantify current revenue loss (spend 1 week with sales to calculate actual monthly deal loss rate and average deal size); (2) If monthly loss exceeds $150K, proceed with agency at $200K budget with penalty clauses (10% refund per week of delay beyond 4 months, 25% refund if crash rate >0."
"Approve agency engagement with structured decision framework: (1) Quantify current revenue loss (spend 1 week with sales to calculate actual monthly deal loss rate and average deal size); (2) If monthly loss exceeds $150K, proceed with agency at $200K budget with penalty clauses (10% refund per week of delay beyond 4 months, 25% refund if crash rate >0.5% at launch); (3) Define 12-month success gate: if mobile drives >25% of new customer acquisition OR >30% of feature requests, transition to in-house team (budget $400K for 2 engineers + manager). If below thresholds, maintain agency relationship but renegotiate to $120K annual maintenance."
Amara Osei "Conditional approval for agency approach with mandatory 2-week discovery phase first."
"Conditional approval for agency approach with mandatory 2-week discovery phase first."
"Conditional approval for agency approach with mandatory 2-week discovery phase first. Spend $10K-15K to: (1) quantify actual monthly revenue loss from lack of mobile app (analyze lost deals, conversion rates, competitor wins); (2) define success metrics (revenue recovery target, quality thresholds, 12-month transition criteria); (3) draft agency contract with penalty clauses for delays, quality gates with refund triggers, and knowledge transfer requirements. Only proceed with full agency engagement if revenue loss exceeds $100K/month AND agency accepts performance-based contract terms."
Recommendation
Executive Summary
The deliberations converged on a validation-first approach that addresses the core flaw in both pure strategies: neither ensures customers will actually pay by Month 8. Sub-problem 1 emphasized that without customer validation, an agency launching Month 4 could generate zero revenue by Month 8, burning £80-120k with no team to iterate. Sub-problem 2 stressed the need for quantified business impact before vendor engagement, warning that 60% of software projects miss deadlines and vendor dependency is a real risk. The recommended approach spends £15-25k over 6 weeks to test actual customer purchase intent (10+ paid pilots or £15k+ pre-sales). If validation succeeds, proceed to agency with performance-based contracts that transfer timeline risk to the vendor and preserve optionality through milestone-based payments. Hire one senior mobile engineer during agency work to build internal capability and prevent indefinite dependency. At Month 5, after agency delivery, make a data-driven decision to either hire a second developer (if pilots convert strongly) or preserve cash (if pilots struggle). This strategy balances Maria's financial risk management with Amara's emphasis on iteration capacity, while protecting you from the most likely failure mode: building a product nobody wants.
Recommendation
Run a 6-week customer validation sprint (£15-25k) to confirm demand before committing to either approach. If validation succeeds (10+ paid pilots or £15k+ pre-sales), hire an agency with performance-based contracts and milestone payments. If validation fails, shut down at £40-55k total spend instead of wasting £100-130k on an unproven product.
How to actually do this
Validation sprint requires active sales involvement: your sales team must conduct customer interviews and negotiate pilot commitments. This cannot be delegated to marketing. Allocate 15-20 hours per week from sales for Months 1-2.,Agency selection depends entirely on validation success. If validation fails, stop immediately—do not proceed to RFP. This requires discipline to kill the project when evidence is weak, even if competitive pressure feels urgent.,Runway monitoring is non-negotiable. Track cumulative spend monthly against plan. If Month 2 validation spend exceeds £30k, kill the project. Scope creep during validation is a warning sign of poor execution discipline.