"How much does it cost?" is the first question almost everyone asks when they look at web push, and it is also the hardest one to answer from a pricing page alone. Two platforms showing the same big number can end up costing very different amounts by the end of the year, because they count different things, put their limits in different places and charge for different features.
This guide does not compare price lists. It explains how the pricing models you will run into actually work, where the costs you do not see at first tend to hide, and how to make a realistic estimate of what you will spend, on your site, with your numbers. At the end you will find Natom's prices and three examples of picking a plan, but the method works for any service you are evaluating.
If you are just getting started and want to understand how notifications work first, read what web push notifications are and come back here.
Why web push pricing is not obvious
Technically, a single web push costs the sender very little. The message travels over the browsers' own delivery services (Firebase Cloud Messaging from Google for Chrome and Android, Apple's infrastructure for Safari), and the platform you use acts as the control room: it collects subscriptions, stores tokens, prepares sends and measures results.
So what you pay for is not really the individual message. It is the service around it: the dashboard, room for your subscribers, analytics, tools for your editorial team, support. Every vendor picks a different unit to charge you for that service, and the unit, more than the headline number, is what decides your bill.
The pricing models you will meet
Pay per subscriber
This is the most common model. You pay according to how many subscribers you have, usually in tiers: one price up to a threshold, the next plan above it. The big advantage is predictability. You know your spend in advance and you can send as many pushes as you like without doing maths every time.
The question to ask with this model is simple: which subscribers count? Everyone you have ever collected, or only the ones who can still be reached? After a couple of years the gap can be large. More on that in the hidden costs section.
Pay per send
Here you pay for how many messages you send, often in bundles of thousands of delivered notifications per month. It sounds fair, but for a site that publishes a lot it is the hardest model to forecast: every extra push to your whole list shows up on the invoice. Newsrooms end up asking "is this story worth the cost of sending it?", which is the wrong question to be asking on a news desk.
Pay per feature
The entry price is low, but features unlock in steps: scheduling on one plan, audiences on another, detailed analytics on a third. That is not wrong in itself, but it means reading the table carefully. The plan you really need is the one that contains the features you will use every day, not the one with the lowest price.
Free plans with limits
Almost everyone offers a free starting point. It is fine for trying things out, and for small sites it can be enough for a long time. But a free plan always has a limit somewhere: on subscribers, on sends, on features, or on something less visible. Knowing where the limit sits matters more than the fact that it is free.
The hidden costs
This is the part people tend to discover later, when switching is already a hassle.
Dead subscribers still counted. A subscriber who uninstalled their browser, changed phone or cleared their browsing data will never receive anything again. If the platform keeps counting them, you keep paying for them. Over time your list grows on paper and your tier goes up, while the number of people actually receiving notifications does not grow at the same pace. Always ask whether there is a regular cleanup of invalid subscribers, and which number your price is based on.
Basic features behind a paywall. Scheduling a push for tomorrow morning, sending only to people who read the sports section, seeing how many clicked: for a publisher these are everyday tasks. If they live on a higher plan, the real price is the higher plan's price.
Hard limits with no warning. Some plans simply stop working overnight when you hit the cap: pushes stop going out, or new subscriptions are refused without any notice. The day that happens is often the day of a big story, because that is exactly when your list grows fastest. A limit that cuts you off without warning is not a price, it is a risk.
Advertising as the price of free. On some free plans you do not pay in money, you pay with your audience: promotional messages shown alongside or instead of your pushes, or your readers' data used for other purposes. Read the terms of a free plan as carefully as the paid ones. If it is not clear what the vendor gets out of your free plan, the answer is usually you.
How to estimate your own cost
To know what you will spend you need three numbers: how many subscribers you will have, how fast they grow, and what each subscriber costs on the plan you would pick.
How many subscribers you will have
Start with your site's monthly visitors. Only some of them will see the prompt, and only some of those will say yes. The opt-in rate depends heavily on the kind of site, the audience, and how and when you ask for permission. There is no right number to copy from somewhere else.
Here is a worked example, just to show the calculation. Say your site gets 50,000 unique visitors a month and, as an assumption, 1% of them subscribe. That is 500 new subscribers a month. This is not a benchmark, it is a hypothesis: your rate could be higher or lower, and you will only find out by measuring how often people accept the prompt during the first few weeks.
How fast they grow
Subscribers do not add up forever. Every month some leave: people change devices, clear their browsers, unsubscribe. In our example, with 500 new ones a month, after a year you will not have 6,000 active subscribers but somewhat fewer. For a cautious estimate, work out gross growth and then subtract a share for losses. It is better to overestimate losses a little than to land in a higher tier by surprise.
Allow for spikes too. One article that does really well can bring in a month's worth of subscribers in a single day. That is why it pays to know what happens when you go over your plan's cap.
Cost per subscriber
Once you have a twelve-month subscriber estimate, divide the plan price by the number of subscribers it covers. That gives you the cost per subscriber per month, which is the most useful figure for comparing plans, even across different vendors. It usually drops as you move up the tiers, but only if you actually fill the plan: paying for a 50,000-subscriber tier while you have 12,000 makes each subscriber four times as expensive.
The indirect costs
The subscription fee is not your only expense. Some costs never appear on an invoice.
Your editorial team's time. Every push has to be written, scheduled and targeted. If the platform makes you copy the title and image of every article by hand, that time adds up every single day. Tools like sending from the post editor, an RSS feed that prepares drafts, or recurring pushes cut this cost more than any discount on the price list.
A developer. The basic setup is simple (a snippet plus one file in your site's root), but customisation, integrations with other systems and configuration problems need someone who knows their way around. On WordPress, a plugin that does everything by itself brings this close to zero. With a custom-built site, budget a few hours of technical work.
Migration. If you already use another service, switching has a cost: time to install the new one, time to check everything works, and above all handling your existing subscribers. The notification permission belongs to your domain, not to the vendor, but subscribers from the old service need to be collected again or migrated carefully. Before deciding, read the migration guide to see what it means in your case.
If you are still weighing up several services, there is a full checklist in how to choose a web push platform.
Natom's prices
Natom uses the per-subscriber model, with unlimited pushes on every plan: you pay for how many subscribers you have, not for how many notifications you send. Prices exclude VAT.
- Free, €0: up to 1,000 subscribers, 1 site, unlimited pushes, full analytics. No card required.
- Starter, €19 a month: up to 10,000 subscribers, 1 site, automatic RSS, scheduling.
- Pro, €59 a month: up to 50,000 subscribers, 3 sites, cross-domain pushes (the same push on several sites), custom audiences.
- Business, €149 a month: up to 150,000 subscribers, 10 sites, API, priority support.
- Enterprise: tailored, beyond 150,000 subscribers, with in-between limits and monthly or yearly pricing.
Two details deal directly with the hidden costs above. First, going over the limit never cuts you off overnight. You get an email notice and 5 days in which everything works; only then do sending and new subscriptions pause until you move to a suitable plan, and you keep every subscriber. Second, subscribers are rechecked and the ones that no longer show any sign of life are removed, so the count you pay for is your active subscribers.
The WordPress plugin is included on every plan, Free as well.
Three examples of picking a plan
The numbers below are made-up examples to show the reasoning, not real data from existing sites.
A small blog
Take a cooking blog with 15,000 visitors a month and, as an assumption, a 1% opt-in rate: 150 new subscribers a month. For the first few months the Free plan is more than enough: 1,000 subscribers, unlimited pushes, full analytics. At that pace, you get close to the cap towards the end of the first year. At that point Starter at €19 a month gives you room up to 10,000 subscribers and adds automatic RSS, handy if you do not want to write every push by hand. Cost per subscriber is high at first (you pay for 10,000 places and use just over 1,000), and falls as the blog grows.
A mid-sized news site
Take a local news site with 600,000 visitors a month, 30,000 subscribers already collected with another service, and dozens of articles a day. Pay-per-send would be a problem here: with several pushes a day the bill would be impossible to predict. The Pro plan at €59 a month covers up to 50,000 subscribers, leaves room to grow, and adds custom audiences, which for a newsroom with different sections (local news, sport, business) are the way to stop sending everything to everyone. If the publisher also runs a second title, Pro's 3 sites cover both.
For newsroom-specific strategies there is a dedicated piece on web push for publishers.
A publishing group with several sites
Take a group with six niche sites that together reach 120,000 subscribers. They need more than 3 sites, so Pro is not enough: the Business plan at €149 a month covers up to 10 sites and 150,000 subscribers, and includes the API to connect pushes to the in-house publishing system, plus priority support. If growth takes them past 150,000 subscribers, the next step is Enterprise, with in-between limits and an agreed price, so there is no forced jump into a much bigger tier overnight.
Monthly or yearly?
Paying yearly costs ten months instead of twelve: two months free, roughly 17% less. In figures, excluding VAT:
- Starter: €190 a year instead of €228 (about €15.83 a month).
- Pro: €590 a year instead of €708 (about €49.17 a month).
- Business: €1,490 a year instead of €1,788 (about €124.17 a month).
Monthly makes sense when you do not yet know how fast you will grow: at the start, during a trial, or if you expect to change plan within a few months. Yearly makes sense once you have a few months of data, you know which tier you will sit in, and push has become a stable channel for your site. A sensible approach is to start on Free or monthly, measure for two or three months, and switch to yearly once your estimate is solid.
Frequently asked questions
Are pushes really unlimited? Yes, on every Natom plan, Free included. The price depends on subscriber count, not on how many sends you make. Common sense still applies: sending too many pushes makes readers unsubscribe, whatever it costs.
What happens if I go over my plan's subscriber limit? You get an email notice and everything keeps working normally for 5 days. If you do not change plan in that time, sending and new subscriptions pause until you do; you keep every subscriber and everything resumes once your plan fits. No surprise on your invoice: you decide when to change plan.
Do I pay for subscribers who no longer receive notifications? Subscriptions get rechecked, and the ones that can no longer be reached (uninstalled browsers, cleared data) are removed. The count you pay for is your active subscribers.
Is there a catch with the free plan? The Free plan has a clear limit: 1,000 subscribers and 1 site, with unlimited pushes and full analytics. As for reader data, Natom does not track your readers' browsing and does no profiling: it stores the subscription token, browser, operating system and device, country, and the topics the reader chose.
Do prices include VAT? No, all the prices listed exclude VAT.
Will web push make me more money? Nobody can promise that. Push is a channel for bringing readers back to your site: results depend on your content, how often you send and how you use it. That is why it makes sense to start small, measure, and move up a plan only when the numbers justify it.