Packaging is deciding what goes into each plan: which features and limits every customer gets, which are held back to encourage upgrades, which are sold as add-ons, and which are left out. Sort every feature into four buckets (core, upgrade drivers, add-ons, leave out), fence tiers with your value metric, and only bundle a new feature into a plan when most customers in that tier will use it.
Key takeaways
- Packaging is what's in each plan. It shapes buying decisions as much as the price.
- Sort every feature into four buckets: core, upgrade drivers, add-ons and leave out.
- Hold back the features that make customers upgrade. Don't give everything away in the entry tier.
- In 2026, the AI question is bundle or add-on. Slack and Notion bundled AI into the plan and raised the price.
- Shrinking what a plan includes without warning causes more backlash than raising its price.
In 2016, GoPro launched its first subscription. For $4.99 a month, camera owners got unlimited cloud storage.
Not a bad deal. GoPro cameras shoot such high-quality footage that storage is a constant headache.
But look at what the package had become by the time I wrote about it in The Subscription Playbook. For $49.99 a year, subscribers got unlimited cloud storage, premium editing tools and a soundtrack library, exclusive deals, up to 50% off accessories and mounts, premium support, and guaranteed camera replacement (up to two cameras a year, basically an unlimited extended warranty).
The price per year went down. The package went way up.
That's packaging. GoPro didn't win subscribers by fiddling with the number. It won them by changing what the number bought. By May 2022 it had 1.74 million subscribers. And when camera sales fell sharply after that, subscription and service revenue held at about $106 million in 2025, roughly flat on the year before, while hardware revenue kept dropping.
So, how do you decide what goes in your package? And what goes in each tier?
What Is Packaging in Pricing?
Packaging is deciding what a customer gets in each plan. It covers:
- Features: what the product or service can do
- Limits: how much of your value metric is included (users, visits, credits, orders)
- Service levels: support, response times, a named contact
- Terms: monthly or annual, contract length, cancellation
Pricing is the number on the tag. Packaging is everything behind it.
Most businesses spend weeks agonizing over whether to charge $49 or $59, then throw features into tiers in an afternoon. It should be the other way round. If the package is right, the price is much easier to set and much easier to defend.
The Four Buckets: Sorting Every Feature
Here's the method I use. Take every feature, service and benefit you offer and put it into one of four buckets.
| Bucket | What it is | Where it goes | Example |
|---|---|---|---|
| 1. Core | The reason anyone buys. Most customers need it. | Every paid tier | Cloud storage in GoPro's plan |
| 2. Upgrade drivers | Things that matter more as customers grow or get serious | Higher tiers only | Extra users, advanced reports, integrations, priority support |
| 3. Add-ons | Valuable to a few, costly to give everyone | Sold separately, or in the top tier | Extra AI credits, onboarding packages, premium data |
| 4. Leave out | Nobody would pay for it, or it adds clutter | Not on the pricing page (maybe not in the product) | Features used by under 5% of customers |
Then follow three rules.
Rule 1: Never hold back the core. If the entry tier can't deliver the main result, customers feel cheated and the brand suffers.
Rule 2: Hold back the upgrade drivers. This is how you grow revenue from existing customers. In the book's land-and-expand chapter I said "you want to hold back on offering everything in free or early subscription tiers." That's still the most common packaging mistake I see.
Rule 3: Be careful with add-ons. Each add-on is another decision for the buyer and another line on the invoice. A few are fine. Ten is a menu nobody reads.
Worked Example: Packaging an IT Support Subscription
Let's package a managed IT service for small offices. (This is an illustrative example.) The value metric is devices supported. Here's how the features sort:
- Core: helpdesk support, security updates, antivirus, monthly health report
- Upgrade drivers: after-hours support, faster response times, backup monitoring, a quarterly strategy review, staff security training
- Add-ons: new-device setup, a one-off security audit, AI assistant licenses
- Leave out: the long list of tools the team uses behind the scenes (clients don't buy tools, they buy results)
Now the tier grid:
| Essentials | Managed | Secure | |
|---|---|---|---|
| Price (per device, per month, ex GST) | $39 | $69 | $99 |
| Devices covered | Up to 10 | Up to 50 | No limit |
| Helpdesk support (business hours) | ✓ | ✓ | ✓ |
| Security updates and antivirus | ✓ | ✓ | ✓ |
| Monthly health report | ✓ | ✓ | ✓ |
| Response time | Next business day | 4 hours | 1 hour |
| Backup monitoring | ✓ | ✓ | |
| After-hours support | ✓ | ||
| Quarterly strategy review | ✓ | ✓ | |
| Staff security training | ✓ | ||
| New-device setup | Add-on | Add-on | Included |
Look at what's happening. Every tier gets the core. The jump from Essentials to Managed is driven by response time, backups and a review, which matter more as a business grows. The jump to Secure is about risk: after-hours cover and training.
Notice also what's not on the grid: no list of 40 software tools. The customer doesn't care which tool you use. They care that their computers work.
There's more on this model in our guide to IT and MSP subscriptions, and on bundling services in general on the bundled services model.
Lay out your packages
Drop your features into three tiers, set the limits and export a clean comparison table with the free Pricing Table Builder.
Bundle or Add-On? The AI Packaging Question
The biggest packaging question of 2025 and 2026 has been what to do with AI features.
There were two schools of thought.
Sell AI as an add-on. Microsoft has defended this route. It launched a $21-a-month Copilot Business add-on for smaller businesses on 1 December 2025, with an $18 promotional price extended to 31 December 2026.
Bundle AI into the plan and raise the price. Slack did this in June 2025, folding AI into its Business+ plan and raising it from $12.50 to $15 per user per month, retiring the separate AI add-on. Notion retired its $10 AI add-on in May 2025 and bundled AI into its $20 Business plan.
Then a third option emerged: bundle an allowance, sell the rest. Atlassian includes 25, 70 or 150 Rovo AI credits per user per month depending on the plan, pooled across the site. HubSpot includes 500, 3,000 or 5,000 credits a month by edition, with more at $10 per 1,000.
Here's how I'd decide:
| Bundle it into the plan when… | Sell it as an add-on when… |
|---|---|
| Most customers in that tier will use it | Only some customers want it |
| It makes the core product better | It's a separate job or product |
| It's cheap to deliver, or you can cap it | Each use costs you real money |
| You want to justify a price rise with visible value | You want to test demand before committing |
If you bundle, the AI in SaaS pricing guide covers how to size the allowance. Read the Slack case study to see how the bundle-and-raise move played out.
The Packaging Change That Angers Customers Most
Here's a pattern I want you to remember.
Customers forgive an honest price rise faster than a quiet cut to what's included.
The 2025 to 2026 AI pricing backlash came mostly from packaging, not price. GitHub Copilot's move to credits drew the line, "You will get less, but pay the same price." Canva cut its free AI allowance from 200 uses to 20 in September 2026. Suno introduced download caps in September 2026 and applied them to songs already created.
In each case, the headline price barely moved. What moved was the package. And customers noticed.
If you need to change a package:
- Tell people early, and show them what it means for their own account.
- Grandfather existing customers for a period where you can.
- Add something when you take something away. A new feature, a longer allowance, a better rate on annual.
And when you do raise prices, the price increase guide walks through how to do it without a wave of cancellations.
A Word on "Unlimited"
"Unlimited" is the most powerful word in packaging, and the most dangerous.
It works when usage is naturally capped. Nobody washes their car 40 times a month, which is why unlimited car wash clubs work so well. It fails when your cost rises with every use. Pret scrapped free drinks in its UK club, and Panera dropped the word "unlimited" from its drink subscription in August 2026, capping it at 30 drinks a month.
In Australia, there's a legal side too. The ACCC issued infringement notices to Amaysim and Lycamobile over allegedly misrepresenting mobile plans as "unlimited" in social media ads. Amaysim paid $126,000 and Lycamobile $12,600 in penalties (ACCC). If your package says "unlimited" but has a fair-use limit, say so clearly and prominently. More on the rules in the Australian Consumer Law guide. This is general information as at October 2026, not legal advice.
My advice: say "a coffee a day", "up to 30 visits", or "unlimited users, fair use applies." Clear limits are generous enough and honest enough.
When to Rethink Your Packaging
Review your packages at least once a year, and straight away if you see any of these signs:
- Almost everyone is on the entry tier. You're giving too much away at the bottom.
- Almost nobody is on the top tier. It's priced too high, or it doesn't contain what big customers want.
- Customers keep asking for one feature from a higher tier. Consider making it an add-on, or moving it down.
- Sales calls are spent explaining the grid. It's too complicated.
- Your costs have changed. Especially if you've added AI features with real running costs.
What to Do Next
- List every feature, service and benefit you offer. All of them.
- Sort them into the four buckets. Ask five customers to help you sort if you're unsure, or run their reviews and emails through the Customer Pain-Point AnalyserPaste reviews, emails or tickets and get pain points and jobs-to-be-done.Open the free tool →.
- Pick your value metric as the main fence between tiers. See value metrics.
- Build three tiers, core in all of them, upgrade drivers stepping up, two or three add-ons at most. The good-better-best guide shows how to space the prices.
- Lay out the grid in the Pricing Table BuilderDesign a good-better-best pricing table and export it.Open the free tool → and test it on someone who's never seen your product. If they can pick a tier in ten seconds, you're done.
If your tiers are based on usage bands rather than features, the tiered pricing guide has the maths. For the full set of decisions, head back to the subscription pricing strategy hub.
Frequently asked questions
What's the difference between pricing and packaging?
Pricing is the amount you charge. Packaging is what the customer gets for it: the features, limits, service levels and terms in each plan. They work together. A great price on a confusing package won't sell, and a clear package makes the price easier to justify. Most companies spend weeks on the number and an afternoon on the package, when it should be the other way round.
Should I put my best feature in the top tier?
Only if it's something your biggest customers need and smaller customers don't. If your best feature is what makes the product worth buying for everyone, it belongs in every paid tier. Holding back the core value makes the entry tier feel broken. Hold back the features that matter more as customers grow, such as more users, advanced reporting, integrations or priority support.
How many features should be on a pricing page?
Fewer than you think. Each tier needs a short list of the five to eight things that matter for the buying decision, starting with the value metric limit. Put the full feature comparison lower down the page, or on a separate page, for the buyers who want it. Lead with differences, not with everything that's the same across tiers.
How do I find out which features customers value most?
Ask them, and watch what they do. A short survey that asks customers to rate features as must-have, nice-to-have or don't care, combined with usage data, will sort most features quickly. Reviews, support tickets and cancellation reasons are also gold. The Customer Pain-Point Analyser can pull themes out of a pile of reviews or emails in minutes.
Can I change what's in a plan for existing customers?
You can, but tread carefully. Removing something customers already pay for feels like a price rise, and often causes more anger than an honest increase. Give plenty of notice, explain why, and consider keeping existing customers on the old package for a period. In Australia, check that your terms allow the change and that it isn't unfair. Get advice if you're unsure.
Get both books free
The Subscription Playbook (ebook + audiobook) and Feed a Starving Crowd, instantly. In Australia? Claim the free hardcover.
Robert Coorey, MBA
Robert is the author of The Subscription Playbook and Feed a Starving Crowd and a Director at BPAG Accountants in Sydney. More about Robert.
Sources
- GoPro: Q4 and full-year 2025 results (5 Mar 2026)
- Slack: June 2025 pricing and packaging announcement
- Atlassian Support: Rovo usage limits
- Visual Studio Magazine: Devs sound off on usage-based Copilot pricing change (27 Apr 2026)
- Suno: Updates to terms of service (2026)
- ACCC: Social media promotions ('unlimited' claims case study)
Facts and prices checked 27 October 2026. Spotted something out of date? Tell us.