Hosting is one of the most leveraged decisions a WordPress agency makes for long-term profitability. The list-price difference between hosting tiers is small. The agency operational time difference is large. The recurring-revenue opportunity from hosting markup is real. The risk of operational debt is real too. After running WordPress hosting across every tier on agency client sites, here is the honest math on how hosting choices feed into agency profitability over five and ten year horizons.
Jump to:
- The three profitability levers in agency hosting
- Lever 1: hosting as recurring revenue with markup
- Lever 2: agency operational time per client
- Lever 3: client retention through reliability
- The per-client profit comparison
- Five-year vs ten-year horizons
- Where agencies lose money on hosting decisions
- The recommendation by agency stage
The three profitability levers in agency hosting
The hosting decision affects agency profitability through three independent levers:
- Recurring revenue from hosting markup. Resell hosting at a higher price than you pay; pocket the difference monthly.
- Agency operational time per client. Hours per month spent on hosting-related work; this is a real cost at your blended rate.
- Client retention through reliability. Sites that go down lose clients. Sites that stay up retain them.
Each lever has different dynamics across the hosting tiers. The right hosting choice optimizes the combined effect.
Lever 1: hosting as recurring revenue with markup
Most agencies resell hosting at a 30-60% markup over their cost. The markup is justified by:
- The agency handles the relationship with the hosting provider.
- The agency provides the value-add layer (monitoring, backups, updates, support).
- The agency takes the operational risk if the hosting fails.
Per-tier markup math (per site per month):
| Tier | Agency cost | Sold to client | Markup |
|---|---|---|---|
| Shared (Bluehost-tier) | $10 | $25 | $15 |
| Managed WordPress (Cloudways) | $25 | $50 | $25 |
| Managed WordPress (Kinsta) | $35 | $75 | $40 |
| Managed VPS shared across 10 sites | $5 | $50 | $45 |
| Self-managed VPS shared across 20 sites | $2 | $50 | $48 |
The markup looks best on self-hosted setups because the per-site cost drops to near-zero. But this ignores the operational time cost, which is Lever 2.
The agency partner programs from Kinsta and WP Engine offer discounted pricing for agencies that bring in multiple clients, plus referral revenue. Real money for agencies with 10+ client sites on the same host.
Lever 2: agency operational time per client
The hours per month per site that the agency spends on hosting-related work:
| Tier | Hours/month/site | Cost at $150/hr |
|---|---|---|
| Shared | 1-3 hours | $150-450 |
| Managed WordPress | 0.2-0.5 hours | $30-75 |
| Managed VPS (multi-tenant) | 1-2 hours per VPS (amortized across sites: 0.1-0.2 per site) | $15-30 per site |
| Self-managed VPS (multi-tenant) | 2-4 hours per VPS (amortized: 0.1-0.2 per site) | $15-30 per site |
Self-hosting wins on amortized per-site time WHEN the VPS is well-utilized (10-20 sites per VPS) AND the operational discipline is in place. Without that, per-site time balloons because incident response is per-site even when routine work is amortized.
The actual operational cost dominates the hosting list price for shared hosting. A "cheap" shared host costs the agency $150-450/month per client in time, regardless of what the hosting plan costs.
Lever 3: client retention through reliability
This is the hardest to quantify but the most important long-term. Clients leave agencies that cannot keep their site up. They stay with agencies that "just work."
Estimated annual churn impact by hosting tier (loose estimates, not rigorous):
| Tier | Reliability-driven churn | Lifetime value impact |
|---|---|---|
| Shared | 10-20% extra churn from hosting issues | Significant |
| Managed WordPress | 1-3% extra churn | Small |
| Managed VPS (well-run) | 2-5% extra churn | Modest |
| Self-managed VPS (well-run) | 2-5% extra churn | Modest |
| Self-managed VPS (badly run) | 15-30% extra churn | Catastrophic |
The takeaway: well-run hosting (managed or self-hosted with discipline) has comparable retention. Poorly-run hosting (cheap shared OR badly-run self-hosted) churns clients aggressively.
The per-client profit comparison
Putting it all together. Monthly profit per client site from hosting, by tier:
| Tier | Markup | Operational cost | Net profit/month |
|---|---|---|---|
| Shared | +$15 | -$300 (avg) | -$285 |
| Managed WordPress (Cloudways) | +$25 | -$50 | -$25 |
| Managed WordPress (Kinsta) | +$40 | -$50 | -$10 |
| Managed VPS (10 sites) | +$45 | -$25 | +$20 |
| Self-managed VPS (20 sites) | +$48 | -$25 | +$23 |
The numbers show:
- Shared hosting actively loses the agency money. The operational time cost is larger than the markup. Agencies that put clients on shared hosting are subsidizing those clients.
- Managed WordPress is roughly break-even on hosting alone. The value is in retention, not in direct margin.
- Multi-tenant VPS (managed or self-managed) is the only tier that meaningfully profits on hosting. And only when amortized across enough sites.
Crucially: the value of managed WordPress is NOT in hosting margin. It is in the operational simplicity that frees up agency time for billable client work. The "loss" of $25/month per site on managed WordPress is more than recovered by the 5-10 extra billable hours per month the team gets back across the portfolio.
Five-year vs ten-year horizons
Hosting decisions compound over time.
Five-year horizon for a 20-site agency:
- Shared hosting: Loss of ~$70k in operational overhead, 20-40% of clients churned out due to issues.
- Managed WordPress: Roughly break-even on hosting; high retention; agency capacity preserved for billable work.
- Multi-tenant self-hosted: Modest hosting profit ($5k-10k); high operational learning curve; risk of one bad incident wiping out the gain.
Ten-year horizon:
- Shared hosting: Agencies that started here are usually consolidated or out of business.
- Managed WordPress: Stable, predictable, easy to scale.
- Multi-tenant self-hosted: Profitable when done right; large operational team or burn-out outcome when done wrong.
The horizons matter because hosting is a recurring decision. The choice you make in year 1 compounds across every subsequent client.
Where agencies lose money on hosting decisions
The patterns I have seen kill agency profitability:
- Putting all clients on shared hosting "because it is what they can afford." The agency eats the operational cost; the client never knows; the agency burns out.
- Self-hosting before having the sysadmin skill in-house. Multiple agencies have done this and rolled back after a serious incident.
- Not bundling hosting into client retainers. Selling the site and then leaving the client to find their own hosting means the agency loses the recurring revenue and the operational visibility.
- Switching hosting providers reactively. Every migration costs real time; pick a host and stick with it unless the host gives you a clear reason to leave.
- Underpricing the hosting line item. Clients see $30/month on the invoice and balk; agencies adjust down. The right move is to bundle hosting into a "care plan" line item ($200-400/month) where the hosting is invisible inside the value-add.
For the broader agency-stack reasoning that hosting fits into, see The Exact Stack I would Use to Run a Small WordPress Agency Today. For the foundational agency-business advice, see How to Start a WordPress Agency in 2026 Without Burning Yourself Out.
The recommendation by agency stage
- Solo freelancer (1-5 client sites): Managed WordPress (Cloudways small plan or single-site Kinsta). Skip the operational overhead entirely.
- Small agency (5-20 sites): Managed WordPress with the agency partner programs at Kinsta, WP Engine, or Cloudways. Negotiate volume pricing.
- Growing agency (20-50 sites): Managed WordPress still; revisit self-hosting only if your team gains real sysadmin skill.
- Scaled agency (50+ sites) with sysadmin skill: Multi-tenant self-hosted VPS for tier-1 marketing sites; managed WordPress for complex workloads.
- Scaled agency (50+ sites) without sysadmin skill: Stay on managed WordPress. The capital cost of hiring a sysadmin to enable self-hosting is hard to recoup.
The pattern: managed WordPress is the right answer for most agencies most of the time. The exceptions (large agencies with strong sysadmin teams, specific workloads that need custom config) are real but small.
For the operational-decision details, see A WordPress Hosting Decision Tree for Agencies, Managed WordPress Hosting vs VPS for Agencies, and When Agencies Should Self-Host WordPress. The complete WordPress cluster covers every adjacent decision.
Sources
Authoritative references this article was fact-checked against.
- Kinsta Agency Partner Program (Kinsta)kinsta.com
- WP Engine Agency Partner Program (WP Engine)wpengine.com
