What It Costs to Run Eleven Products Solo: Our Actual Monthly Bill (2026)

Quick verdict

Our monthly operating bill for running roughly eleven products comes to about £220. That works out to around £20 per product per month for hosting, email, productivity, monitoring, banking, and transactional fees combined.

The bill is dominated by a handful of paid services, a self-hosted backbone that costs nothing in subscriptions but eats VPS resources, two lifetime deals from years ago that are still running, and a careful avoidance of premium tiers we would not actually use.

There are two fee traps in this stack that nobody warned us about, and they each cost real money. Both are below.

This article is the full itemised picture - what we pay for, what we deliberately skip, and the per-product math that makes a lean portfolio operation viable for a solo founder.

The full itemised bill

In rough order of monthly cost, with real prices:

Claude Code (Max plan) - £150/mo - AI coding assistant. Largest single line item.

Hostinger VPS - £17.99/mo - One VPS running all eleven products. Shared infrastructure.

Resend - $20/mo (about £16) - Transactional email across the portfolio. More on the multi-domain trap below.

Business banking - £12.50/mo plus 35p per incoming transaction - The 35p fee bit us in a specific way described below.

Domain renewals - roughly £20-30/mo amortised - About ten domains, mixed TLDs. .com renewals are £8.99/year each; niche TLDs (.gg, .ai) are several times that.

Paddle - 5% plus $0.50 per transaction - Payment processor for the paid product. Variable cost, but the per-transaction floor matters at low sub prices.

Cloudflare - free - Turnstile bot protection on forms. DNS for sites that need it.

GitHub - free - Public repos and CI. Private repos where needed.

Bitwarden - free - Password manager. Free tier is more capable than most paid competitors.

Obsidian - free - Notes and knowledge base. Using a free sync plugin instead of paid Obsidian Sync.

MXRoute - free (lifetime payment years ago) - Business email across the portfolio. The kind of lifetime deal that aged well.

Umami - free - Self-hosted analytics. Runs on the VPS, costs nothing extra.

Uptime Kuma - free - Self-hosted uptime monitoring. Same setup as Umami.

Total recurring: roughly £215-230 per month, plus variable Paddle and banking transaction fees.

The hosting backbone

One Hostinger VPS at £17.99/month runs everything. All eleven products share the same infrastructure - Nginx as the reverse proxy, CloudPanel for site management, Umami and Uptime Kuma running alongside the production sites as Docker containers.

The math here matters. Vercel-per-project pricing on a free tier works for one or two static sites. For a portfolio that includes some app-backed projects, per-project hosting costs would scale linearly. A VPS scales by resource use, not site count - and at our traffic levels, eleven static-export sites barely register on a 4-core VPS.

The trade-off is real. With a VPS you manage Nginx config, you handle Docker yourself, you set up your own backups. If you do not enjoy that work, the cost saving is not worth it. For us, the discipline of running our own server has paid back in portability - we are not locked into any platform.

The Resend multi-domain trap

Resend's free tier covers 3,000 emails per month and up to 100 verified domains in theory, but the practical limit hits sooner. Every Vine in the portfolio is a separate domain. Every domain that sends transactional email needs SPF, DKIM, and DMARC records, plus a verified sender on Resend.

For a single-domain founder, Resend is genuinely free. For a portfolio operator running ten-plus domains, you trip into the paid tier even at moderate sending volumes. We pay $20/month - the entry-level paid plan - because the number of verified senders we need is what pushes us over.

This is the kind of cost detail nobody warns you about when you start. The "free tier" language is technically accurate but only applies if your business model fits within the free tier's assumptions. Multi-domain operations have different free-tier ceilings than the marketing suggests.

Worth knowing before you commit to a portfolio approach: every domain you add has compounding email infrastructure cost, even if each individual domain sends very little.

The Paddle fee math at low sub prices

Paddle charges 5% plus $0.50 per transaction. On a $50 annual purchase, that is $3.00 - a 6% effective rate. Fair.

On an £8.99 monthly subscription, that is roughly £0.45 plus £0.40 (USD to GBP) - around £0.85 in fees. Effective rate: about 9.5%. Brutal on small recurring subs.

Why pay Paddle's higher fees instead of Stripe's 2.9% plus 30p (about £0.56 on £8.99, effective ~6.2%)? One reason: Paddle is a Merchant of Record. They handle VAT registration, sales tax compliance, and invoicing in every jurisdiction they sell into. Stripe expects you to handle tax compliance yourself.

For a sole trader selling globally, MoR is worth real money. For a UK-only sale, Stripe wins. The choice is not "which has lower fees" - it is "which fee structure matches the actual scope of where you sell."

The lesson: do the math on your actual sub price, not the marketing example. A 5% fee sounds clean. 5% plus a fixed fee on a small recurring sub is not clean. Always run the calculation on your real numbers.

The banking 35p trap

Our business bank charges £12.50/month for the account plus 35p per incoming transaction. This is normal for UK business banking - challenger banks often have similar structures, sometimes worse.

The 35p per incoming bit us in a specific way. Affiliate networks sometimes send tiny sweep payments to clear out micro-balances - PayPal and Amazon Associates both did this on different occasions, each sending a 1p test transaction to our business account.

A 1p incoming payment cost 35p to receive. That is a 34p net loss per transaction, on a payment whose face value was 1p.

It is funny in retrospect. It was less funny when we discovered we were running negative on the affiliate payouts we had been celebrating. The lesson: your bank can cost you money even when affiliates pay you out. If you are running affiliate income, set a minimum payout threshold higher than your bank's incoming-transaction fee.

Worth setting up: in any affiliate dashboard that lets you, raise the minimum payout. Most of them default to absurdly low minimums that benefit no one once banking fees are factored in.

Two lifetime payments that aged well

MXRoute for business email: paid once, years ago, still running. We use it for inboxes across the portfolio. The deal was inexpensive at the time and has effectively paid for itself many times over against what Google Workspace or Fastmail would have cost over the same period.

The principle: when a tool you actually use offers a real lifetime deal at a sensible price, take it. Recurring costs compound. A £80 lifetime deal on something you would otherwise pay £8/month for breaks even at ten months.

The caveat: lifetime deals from companies that disappear are worthless. MXRoute happened to be run by a small operator with a track record. We have also bought lifetime deals on tools that vanished within two years. The track record of the seller matters more than the price.

Some deals that age badly: bundled lifetime deals on AppSumo from companies that pivot, get acquired, or shut down. Some deals that age well: indie-run infrastructure tools with clear hosting costs and a sole operator who has been around for years.

What we deliberately do not pay for

This is the other half of a lean operation. The skipped subscriptions matter as much as the paid ones.

No 1Password. Bitwarden free tier covers what we need. The polish difference is real, but the cost difference compounds.

No Obsidian Sync. A free community sync plugin handles cross-device syncing. Obsidian itself is free for personal and small business use.

No paid analytics. Umami self-hosted gives us what Fathom or Plausible would charge for. The trade-off is we maintain it ourselves.

No paid uptime monitoring. Uptime Kuma self-hosted gives us alerting and dashboards. UptimeRobot free tier would also work for one or two sites; at our scale, self-hosted is more practical.

No newsletter SaaS. We do not have lists big enough to justify Kit, Beehiiv, or Substack subscriptions yet. When we do, we will pay for one.

No GitHub paid. Public repos for open-source work, private repos within free-tier limits. We have not needed the paid plan.

No premium tiers on tools where we use the free tier. Cloudflare free is enough for our purposes. GitHub free is enough. Bitwarden free is enough.

The pattern: pay for tools that solve a real bottleneck. Skip premium tiers that solve nice-to-have problems. Self-host where the maintenance cost is less than the subscription cost over a sensible timeframe.

The self-hosted backbone

Three tools run on the VPS at zero monthly subscription cost: Umami for analytics, Uptime Kuma for monitoring, MXRoute for email (technically on a different server, but the lifetime payment model is similar).

The honest accounting: self-hosting is not free. You pay in VPS capacity, your own time to set things up, the maintenance burden when something needs updating, and the on-call obligation when something breaks at 2am. The subscription cost is replaced by an attention cost.

For us, the trade is worth it because we are running multiple products on the same infrastructure. The marginal cost of adding Umami to a VPS we are already paying for is essentially zero. Setting up Uptime Kuma alongside Umami was an afternoon. The fixed time investment amortises across all eleven products and continues to compound.

For a single-product founder, the math may not work. Setting up a self-hosted stack for one site is overkill - the time investment will likely exceed what a paid Fathom subscription costs over the same period.

Could you start this cheaper?

A hypothetical day-one stack for someone starting today, without the AI coding spend:

Hostinger or Hetzner VPS: £5-15/mo for a small VPS that handles a static site or two.

Cloudflare: free tier covers DNS, CDN, Turnstile.

Resend: free tier up to 100 emails/day for a single domain - plenty for early-stage.

Umami self-hosted on the VPS: free.

GitHub: free.

Bitwarden: free.

Obsidian with free sync plugin: free.

Total entry point: £5-15/month for the operational stack. The AI coding subscription is the largest discretionary cost, and it is genuinely optional - everything we run could have been built without it, just more slowly.

Domains add up depending on how many you register and which TLDs you pick. Stick to .com if you can - £8.99/year is much cheaper than £60-100/year for the trendy alternatives.

The honest takeaway

About £220 per month for a portfolio of eleven products. Roughly £20 per product per month for the operational stack.

A meaningful chunk of that goes to the AI coding subscription, which is the most discretionary line item and the easiest to scale back if cash gets tight. The rest is the actual operational floor - hosting, banking, transactional email, payment processor fees.

The lean discipline matters. Premium tiers on tools we would not fully use, paid SaaS where a self-hosted version works, recurring subscriptions where lifetime deals exist - those are all places we deliberately stayed cheap. Each individual choice is small. They compound.

The fee traps - Paddle on small subs, the bank's 35p per incoming - cost real money when you are not watching. Watch them.

Most founder-cost articles either inflate the bill (to make their consulting look essential) or hide it (to make affiliate-promoted tools look free). Ours is just the bill. Make of it what you will.