Northbeam and Ripplux are not really competing for the same customer. Northbeam is enterprise measurement infrastructure. Ripplux is a measurement tool for an operator who does not have a data team. The clearest way to see the difference is the price floor: as of 2026-07-28, Northbeam's published pricing starts at $1,500 per month. Ripplux starts at $149.
That gap is not arbitrary. It buys real things. Whether it buys things you need depends almost entirely on how much you spend on ads.
What each tool actually does
Northbeam is a multi-touch attribution platform with warehouse-grade ambitions. It ingests large volumes of behavioural data, builds device and identity graphs in-house, models both clicks and deterministic view-through conversions, and offers media mix modelling as an add-on at higher tiers. Its published pricing is based on data volume measured in pageviews rather than on revenue, which tells you what the product is: a data processing system.
Ripplux runs randomised holdout experiments on Meta and Google. It reads Shopify orders and ad spend through official APIs, withholds a campaign from a randomised slice of your audience, and measures the difference in orders. It builds no identity graph, models no customer journey, and produces no configurable dashboard. It returns a causal estimate and a recommendation.
The philosophical split is that Northbeam works to observe more of the journey and model it better, while Ripplux gives up on observing the journey and tests the outcome instead. Both are legitimate. They fail in different directions: modelling degrades quietly as tracking coverage degrades, and experiments simply cannot run when volume is too low.
Feature comparison
| Northbeam | Ripplux | |
|---|---|---|
| Core method | Multi-touch attribution with view-through | Randomised holdout experiments |
| Identity graph | Yes, built in-house | No |
| Media mix modelling | MMM+ as an add-on | Not offered |
| Incrementality testing | Not the core method | The entire product |
| Custom dashboards and reporting | Yes, warehouse-grade | No |
| Channels covered | Many | Meta and Google |
| Published entry price | From $1,500 per month | $149 per month |
| Pricing basis | Data volume in pageviews | Flat per plan |
| Contract | Month-to-month on Starter, annual above | Monthly or annual, cancel any time |
| Entry path | Sales-led on higher tiers | Self-serve from the Shopify App Store |
Pricing, side by side
Northbeam publishes three tiers. Starter is listed as starting at $1,500 per month for businesses under $1.5M in annual media spend and is billed month to month. Professional is custom-priced and aimed at brands spending over $250,000 per month. Enterprise is custom-priced and aimed at over $500,000 per month. Northbeam states that price is informed by data volume measured in pageviews, and that Professional and Enterprise require annual terms with annual or semi-annual billing.
Ripplux is $149 per month for Pro, which includes one holdout experiment per calendar month, and $299 per month for Growth, which includes unlimited experiments. Pricing does not scale with revenue, ad spend, or pageviews.
The ratio that decides this
Measurement tooling is overhead on the budget it governs, and the useful question is what share of the budget it consumes.
At $20,000 a month in ad spend, Northbeam's published floor of $1,500 is 7.5% of the budget. Measurement has to find a lot of waste to justify taking that much off the top. At $200,000 a month, the same $1,500 is under 1%, and if the modelling reallocates even a small share of spend correctly it pays for itself many times over.
Ripplux at $149 is 0.75% of a $20,000 budget and under 0.1% of a $200,000 one. But Ripplux does less. It does not model your full customer journey, it does not cover every channel, and it does not give your analyst a surface to work in.
So the honest framing is not "which is better" but "which overhead ratio matches what you get." Below roughly $50,000 a month in ad spend, a $1,500 floor is hard to justify for most solo operators, and the causal question can be answered directly with experiments. Above roughly $150,000 a month, with a data team in place, Northbeam's modelling surface starts doing work that experiments alone cannot.
When Northbeam is the better choice
If you have a data team, Northbeam is built for them and Ripplux is not. Ripplux deliberately gives you no knobs. That is a feature for an operator and a limitation for an analyst who wants to interrogate the model.
If you need many channels covered, Northbeam covers them. Ripplux runs holdouts on Meta and Google, which covers most Shopify merchants' spend but not a brand running TV, podcast, affiliate, and retail media alongside.
If you need view-through measurement, Northbeam models it explicitly with its own graphs. Ripplux has no view of impressions outside the platform reporting, and a holdout experiment measures total campaign effect rather than decomposing it by touch type.
If you need continuous daily readouts across everything, modelling gives you that and experiments do not. An experiment concludes on its own schedule.
When Ripplux is the better choice
You are the person who runs the campaigns. There is no data team, and there is not going to be one this year. You want to know which spend is not causing sales, and you want the answer without becoming an analyst.
Your ad spend is somewhere between $5,000 and $50,000 a month, which is enough for a holdout experiment to reach a usable confidence interval but not enough to absorb a four-figure measurement bill.
You want the entry path to be an install rather than a sales cycle. Ripplux installs from the Shopify App Store, connects through OAuth, and starts reading history immediately.
What Ripplux will refuse to tell you
A holdout needs volume. Ripplux requires roughly $5,000 in monthly spend on the platform being tested and about 50 conversions per month before a test is worth running. Under those thresholds the confidence interval gets wide enough that the result cannot support a decision, and Ripplux says so rather than reporting a number that looks precise and is not.
If your spend sits below that floor, neither tool on this page is the right purchase yet. That is a real answer, and it is cheaper than either subscription.
Can you run both?
Above roughly $10M in revenue some brands do, with Northbeam as the modelling and reporting layer and Ripplux as the periodic experimental check on it. A model is a hypothesis about how your media works, and a holdout is the closest thing to a test of that hypothesis. Running an experiment against a model's prediction is how you find out whether the model has drifted.
Below that scale, running both is paying twice to answer one question.
Frequently asked
- At what spend level does Northbeam make sense?
- Northbeam's published pricing starts at $1,500 per month for brands under $1.5M in annual media spend, with higher tiers aimed at $250k and $500k in monthly spend. As a rough test, measurement tooling should cost a small fraction of the budget it governs. At $20,000 a month in ad spend, a $1,500 measurement bill is 7.5% of the budget. At $200,000 it is under 1%.
- Does Ripplux replace Northbeam?
- For a brand with a data team, a warehouse, and six-figure monthly spend, no. Northbeam does modelling work that Ripplux does not attempt. For a solo operator spending $5,000 to $50,000 a month, Ripplux answers the causal question directly at a fraction of the price, without a data engineer.
- What does Northbeam do that Ripplux does not?
- Multi-touch attribution across many channels, view-through attribution using its own device and identity graphs, media mix modelling as an add-on, and a customisable warehouse-grade data surface. Ripplux runs holdout experiments on Meta and Google and returns a verdict. It builds no identity graph and no dashboard.
- Is onboarding self-serve?
- Ripplux installs from the Shopify App Store and connects through OAuth, with no sales call. Northbeam's published pricing routes Professional and Enterprise through custom quotes on annual terms, so the entry path runs through their sales team.