Measurement Platforms Full Review
- Type: Multi-Touch Attribution
- Platforms: Web (Desktop) + Offline
- Core Offering:
- Cross platform attribution, incrementality experiments (requires turning off campaigns)
Rockerbox was founded in 2013 as a solution to determine user intent by analyzing browsing behavior. The idea behind the platform was to use recency as the most important parameter to understand intent.
By creating a network of user identification and learning from how recent users that made a purchase behaved vs. users that are potential to do so.
RockerBox measurement is based on identifying users browsing activities over the past 30 minutes, identifying the top sites, articles and topics that users engaged with before and after arriving to your website to effectively put marketing plans.
During 2018, Rockerbox introduced a ”multi-touch” attribution solution, by allowing customers to consolidate marketing reports across web and offline campaign sources to create some sort of a “weight” for each touch point.
During 2020, Rockerbox added the ability to track the incrementality tests you are performing (manually) by logging those tests and trying to make more sense out of the data by connecting the results of the hold out (“blackout”) tests into the Rockerbox dashboard in order to try and make more sense of the data.
Integration / Development Needed
Integration with Rockerbox takes between 2-4 weeks, and requires several conversations to understand our marketing activities and map out which integrations are in order.
Rockerbox offers multiple platform integrations:
- Sending files to Rockerbox for Direct mail, TV and radio postlogs, batch spend, batch OTT, retail conversions
- Reports should include: Timestamp, Network, inventory, media outlet type, feed type, ad name, tape lengh
- Using Webhook or click tracker using pixels
- Integrating call center data (invoca, CallRail)
- Sending aggregated “direct mail” data
- Requires: Mail date, in home date, street, city, state, zip code, country
- Amazon seller data
- OTT log files
Mobile is not supported. Rockerbox only supports an integration using Branch, and as an alternative, requires you to pass data files that include:
- Customer email
- User ID
- Device ID
- Device IP Address
- Order ID
Note: The hooks marked in red are no longer allowed by mobile operating systems without user direct consent to App Tracking Transparency (ATT)
Additional Note: Rockerbox does not support ingesting returns. This means that if a channel is producing fraud sales (which 100% of those get refunded) – RockerBox will not handle this!
Rockerbox does support an integration with Google AdWords – but for every other platform you will need to onboard data, you will need a lot of support from your engineering and BI teams pulling data and connecting to RockerBox via APIs.
Rockerbox is an advanced measurement platform allowing a marketer to create a unique CMO dashboard consolidating data across offline and online campaign activities to provide oversight. The platform is an enterprise platform that requires a heavy investment in resources to integrate and onboard.
Rockerbox came up with 4 stages to onboard:
- Goals and priority use cases
- Implementation and data collection
- Platform setup
- Platform use
The onboarding process can take several months.
The setup part is very manual where you need to input a lot of details over every marketing activity you ran and are running. Rockerbox encourages you to onboard data across all sources for the past 12 months – this includes every promotional email you sent, every podcast sponsorship, every direct mail, TV campaign, and of course, all digital campaign activity you ran.
As Rockerbox works best for websites and offline brands – things like “app updates” and “features” are not considered as relevant data points
Rockerbox offers an SLA providing support only during business hours (9am to 6pm Eastern time) and does not offer any commitment to a response time.
They do have a very thorough documentation area within the website, but unfortunately we could not find a support desk, tickets platform or an on-call number to reach out for quick questions.
While the integration and onboarding was professionally handled, we did expect better support from Rockerbox.
Maybe they should invest in a support desk ?
If you did manage to integrate your data source, including historical data, and manually input everything you’ve done in the past 12 months (including every promotional email you sent, every featuring you got, every sale, every mail, tv campaign, emailer, social post and so on) Rockerbox will visualize the data in a very useful way.
The platform was designed for CMOs more than it was designed to marketing managers.
So – if your role is to operate campaigns, make tactical and strategic decisions while handling a budget – don’t expect useful insights from Rockerbox.
But if your role is to only make decisions about your overall budget, media mix decisions – Rockerbox can provide useful analytics to you and make yourself easier.
Rockerbox does have some discrepancies with ad networks and your own data , but they do work with you and explain where discrepancies come from.
- Discrepancies between ad platforms and Rockerbox
- Rockerbox explains how Google and other self reporting platforms claim attribution based on their own ad serving, vs other ad networks and ad platforms.
- Rockerbox deduplicates the data based on last touch and gives full credit only to the ad network that “touched” the user last.
- Facebook Discrepancies
- Rockerbox specifically addresses Facebook discrepancies, as Facebook claims conversions for both postview, and post click by default, while Rockerbox does not.
- Spend Discrepancies
- To visualize your spend, Rockerbox does use some calculations and triangulation in order to assign a spend figure for activities that may not be correct.
- i.e. if you ran paid Mail campaign (not email – MAIL) and input a one time spend amount, Rockerbox may distribute this spend amongst several channels or one, either proportionally or not. The way Rockerbox ”decides” to split your spend is based on the preliminary onboarding calls you’ll have with them. We found this very confusing.
- Discrepancies between Internal data and Rockerbox
- There are several issues that can cause additional discrepancies:
- Timezone difference
- Difference in implementation
- Difference in user deduplication (i.e. you would consider a customer as a unique user, while rockerbox may identify the same customer as 2 or more users)
- Page load time of the Rockerbox pixel
- Order cancellations
- There are several issues that can cause additional discrepancies:
NOTE: Rockerbox does not ingest cancellations and Refund data. If your product offers users any option to refund / opt out – Rockerbox could be offering a crazy loophole for fraudulent activity (see under Security). You could see thousands of conversions on Rockerbox, which will end as 100% refunded, and Rockerbox will recommend marketing strategies to you based on this false data!
Getting integrated with Rockerbox took a lot of work, and a lot of different reports and APIs we needed to connect.
This did cause strange discrepancies as described in the “accuracy” section.
But the absolute worst potential breach with Rockerbox is their inability to ingest Cancellations and Refund data.
This meant that whenever we faced an ad network that supposably generated good results only to later find out that the results were all fraud and 100% of the “sales” were cancelled – Rockerbox was unable to process this data, and continued providing us with wrong insights and wrong recommendations considering the sales generated by the fraud source as legitimate.
It’s a yes and no for Rockerbox.
When it comes to the data they request for integration, the data requested includes things like Email Address, IP, Physical customer address(!) and many other customer information.
While this is your own 1st party data – sharing this with any provider is very much within the gray area of compliancy.
For web tracking, Rockerbox uses either 3rd or 1st party cookie data.
For mobile, Rockerbox only supports Branch analytics and is incapable of attributing mobile app conversions themselves.
Unbiased vs. Biased
Related to the pricing model and price – Rockerbox is a true saas company that does not charge based on your spend or revenues.
We also found no affiliation between Rockerbox and any ad network or marketing platform.
Rockerbox is a relatively small company with only 35 employees.
Pricing Model & Price
Rockerbox pricing model is a pure SaaS model. You only pay for a license to use the platform and are not charged based on either your media spend, traffic volume, or revenue.
We appreciated this.
The license fees was anywhere from a few thousands of dollars, to over $20,000 depending on the amount of integrations you wish to include, the channels and the complexity of your own setup.
Once you have established this – you don’t need to worry about how your spend it ramping up or down.
Rockerbox does require an annual commitment with no exit clause – which means, that once you made a decision to use the platform – expect to make the resources available for the integration and for paying the enterprise licensing fees.
Rockerbox is a marketing measurement platform allowing the consolidation of data across multiple platforms. These are very much web platforms (i.e. adroll, bing, criteo, outbrain). Campaign tracking requires the use of UTM codes. Onboarding takes between 2-4 weeks as per documentation, but our experience was closer to 3 months to get onboarded.
The attribution part of the platform is not tailored to any advertiser investing in app promotion / advertising.
But with that , as a media mix modeling platform, Rockerbox uses historical data (in aggregation) suggesting the right combination of channels and mediums.
Rockerbox includes an incrementality section, suggesting a series of experiments (turning off campaigns in across TV or specific channels like Facebook and Google) to measure the impact. Rockerbox also uses the baked in “lift” measurement packages offered by some platforms.
Note: Facebook deprecated Lift tests as of 2021
Pros & Cons:
- The dashboard looks really nice and can act as a CMO dashboard
- Data can be onboarded from various channels. Rockerbox works very well as a data visualization tool.
- Path to Conversion is the most valuable feature Rockerbox offers – but this only works on browsers supporting 3rd party cookies.
- Rockerbox does not support attribution for iPhone and iPad apps
Measurement Platforms Review for RockerBox:
|No Dev Needed||★★★★★☆☆☆☆||Privacy Compliant||★★★★☆☆☆☆☆|