Is there a good reason for the QuickSight pricing model?

I am really puzzled by the QuickSight pricing model. It seems to be so not AWS-like. Everything I can think of at AWS is on usage-based billing, sometimes is so granular it’s fascinating.

But with QuickSight there are just 2 plans, where most of the advanced features are locked up in the Enterprise tier, and even then you need to pay minimums per account to get them.

I feel like this is very much contrary to the AWS spirit which also could hinder wider adoption, security and development experience.

AWS recommends having a separate sandbox environment for each developer. Yet, this is not possible with QuickSight, as then each developer needs to have a QuickSight account, with all of the premium features enabled and paid for, this can easily translate to $1,000/dev/mo.

And then on top of that, I have customers who want to use some of the gated features, e.g. paginated reports, but they do not need $500/mo worth of them. So they are not signing up. Maybe they would use $100/mo worth of usage, but now AWS is missing out on this revenue.

As a consultant developer, I took a risky bet on QuickSight early on, and I keep recommending it to our customers, but I do feel like I am being sidelined with these inaccessible premium features that can only be paid for by truly enterprise customers. But at the same time, I do not feel at all that these features are “enterprise”-grade. Rendering a report to PDF programmatically IMO should be table stakes in 2023.

Could someone please explain what is the thinking and strategy behind this pricing?

Is there a future where this will be unbundled?

3 Likes

Hi @m0ltar Thanks for your question. I have reached out internally to find someone who can speak to this.

I second this question. You have such a great pricing model going over Tableau and Qlik, fixing this would make Quicksight a knock out no brainer.

1 Like

Hi all! I have been in contact with one of our Solution Architects who plans to reply to this.

Hi @m0ltar ,

We appreciate your feedback! At AWS we are customer obsessed and work to accommodate any feedback we get.

Regarding the two separate plans, the Standard Edition is meant for individual users who would like to simply create analyses on their data. The Enterprise Edition (vast majority of our customers) is for organizations who need enterprise grade features such as row level security, embedding, and user permissions.

In the Enterprise Edition, Authors are monthly pay as you go. You can also receive a discount per author with an annual commitment (similar to Amazon EC2 reserved instances). The value of this is that you only pay for Authors you have added to your QuickSight account, no need to over index and get locked into a contract. If Authors are not logging in, simply remove them and lower your cost. Author prices increase slightly if you add QuickSight Q. There is a $250/month base for Q covering the extra resources needed to build the natural language and ML models.

Pricing flexibility extends to Readers. Customers can sign up month to month and pay only $0.30/session with a $5 max/month. Meaning, customers only pay for the sessions they use. No need to pre-provision and pay for users that infrequently or never log in. For a further discount, customers can provision capacity (# of sessions) instead of individual users (again, similar to Amazon EC2 reserved instances). These charges increase slightly per user or session if QuickSight Q is enabled. Economies of scale are built into the pricing as customers use more sessions as well.

Amazon QuickSight also has separate pricing for some features, so that customers who do not use them, can save money. Minimums exist to cover infrastructure needed to support these features, however they are largely usage-based. Paginated reports start at $500 a month. However, regular QuickSight dashboards include PDF scheduling at per session cost. Alerts start at $0.50 per 1000 metrics evaluated ($0.0005 per metric evaluated). This is also separated out to provide a large discount from the $0.30 standard session pricing. Both of these features also have economies of scale built into pricing as customers use them more.

SPICE (Superfast Parallel In-memory Calculation Engine) capacity is $0.38/month. This is pay as you go. SPICE capacity can be added or removed at any time. Every QuickSight account includes 10GB extra capacity per author and is pooled per account/region. Datasets using direct query do not count towards your SPICE capacity.

It seems that your concerns are mainly focused on the minimums for various functionality. I would like to discuss that further with you. I will be messaging you shortly to get your contact information.

4 Likes

Great points @m0ltar.
@bergqdou the minimums in particular for paginated reports have been raised across the forum, e.g.

Aside from the fact that I also think that paginated reports isn’t a $500 a month feature (in fact I’m surprised that it’s even considered a premium feature), the fact that it has a significant fixed cost feels out of kilter with AWS in general.

We might dabble with it for a few months and see what our customers think, but we’re not going to put much time and effort to it as we don’t want to carry that cost if our customers don’t see it of real value (and of course as soon as you offer it, even if only one customer actually uses it, you then have a very unhappy customer if you withdraw it). And you can’t pass on the cost to customers at low scale, because it’s a large single fixed cost.

So yeah, setup feels quite wrong at the moment.

1 Like