AD cube
AD cube is an Italian AI suite for digital advertising that brings marketing mix modelling, real-time cross-channel budget optimisation and creative performance analysis together in one platform for high-spend performance marketing teams.
What is AD cube?
AD cube is an AI-powered digital advertising suite built by AD cube S.r.l., a Milan company born out of research at the Politecnico di Milano and incubated by ML Cube. Five co-founders stand behind it, three of them academics working on machine learning and optimisation in economics. The platform itself lives at app.adcube.ai and targets in-house performance teams and agencies handling substantial multi-channel media budgets.
The suite is organised as three linked pillars. AD cube MMM covers planning: it measures how much each channel contributed to past results, lets teams simulate new allocation scenarios without spending, and returns an optimised week-by-week budget plan for the year ahead. Unusually, the vendor claims the model copes with data of any quality, where most marketing mix models demand clean historical series. AD cube Core covers execution: predictive curves forecast campaign revenue and ROI, and online machine learning redistributes the daily budget across campaigns at ad-group level, within limits the advertiser sets. Creative DNA closes the loop on analysis: it auto-tags every element of an image or video ad, from colours, brightness and layout to CTA placement, product packaging and indoor or outdoor setting, runs OCR over text inside creatives, detects prices and discount codes, clusters similar assets and issues recommendations that refresh as new data arrives.
Two add-ons extend the suite. Ads GPT generates, edits and publishes Meta and Google Ads campaigns from a natural-language instruction, with preview and approval before anything goes live; it is still offered through an Early Adopters programme rather than being generally available. Audience Discovery scrapes and analyses social pages and behavioural data to build machine-learning lookalikes, behavioural clusters and interactive interest treemaps.
Twenty-five advertising and analytics platforms connect to it, split between nine Ready to go integrations, among them Google, Meta, TikTok, Amazon Ads, DV360, Search Ads 360 and Google Analytics, and sixteen Premium ones that reach demand-side platforms such as Xandr, Yahoo DSP and Amazon DSP. Onboarding is presented in three steps: connect the account, import campaigns and set the optimisation goal, then let the system optimise on auto-pilot or under custom constraints. No price is published anywhere: the three products are advertised from one million euros of annual media investment and everything goes through a tailored quote.
What it does
- Model the marketing mix and measure how much each channel really contributed to results
- Simulate what-if budget allocation scenarios without spending money
- Produce an optimised week-by-week budget plan for the year ahead
- Reallocate the daily budget automatically across campaigns and channels at ad-group level
- Forecast campaign revenue and ROI through predictive curves with advertiser-defined limits
- Auto-tag, cluster and score ad creatives, then act on the recommendations that follow
- Generate, preview and publish Meta and Google Ads campaigns from a natural-language prompt
When to use AD cube / When not to
A quick filter to help you decide if AD cube is the right fit.
When to use AD cube
- Advertisers spending more than one million euros a year on media, the threshold the vendor advertises for all three products
- In-house performance marketing teams running campaigns across Meta, Google, TikTok, Amazon Ads and demand-side platforms
- E-commerce and retail brands that have to justify how budget is split between channels
- Finance and lead-generation advertisers judged on cost per qualified lead rather than raw volume
- Creative and brand teams that want evidence on which creative elements actually drive results
When not to use AD cube
- Small advertisers and freelancers: every product is advertised from one million euros of annual media investment
- Anyone wanting a self-service purchase, since there is no free plan, no free trial and no published price
- Teams looking for organic social media publishing or community management
- Anyone expecting creative generation: Creative DNA analyses existing assets, it does not produce images or video
- Mobile-first users, as the product is web-only with no app and no browser extension
How to use AD cube
A typical end-to-end flow, from setup to results.
- Book a demo through the Calendly link on the site, or send the contact form on the homepage
- State your annual ad spend bracket and whether you want a demo or a conversation with an expert
- On the pricing page, select the annual advertising investment band that matches your business
- Choose the products you are interested in: AD cube MMM, AD cube Core, Creative DNA, or a combination
- Give your name, company and business email to request the tailored quote, answered within 24 hours
- Sign the Conditions of Use and the Subscription Form, adding a data processing agreement if you need one
- Connect your advertising accounts to the platform, presented as a one-click operation
- Import your campaigns and define the optimisation goal you want the models to pursue
- Let AD cube rebalance spend on auto-pilot, or tune the constraints and limits yourself
- For Creative DNA, connect Meta, Google or Amazon, pick the campaigns to analyse and set your tags
Pros & Cons
Pros
- End-to-end coverage that is rare in this category: annual planning, daily budget optimisation and creative analysis in one platform
- Verifiable academic roots, with research from the Politecnico di Milano, incubation by ML Cube and co-founders who are professors and PhDs in artificial intelligence
- The marketing mix model is claimed to work with data of any quality, where competing models require clean historical series
- Broad platform coverage, with twenty-five connectable advertising and analytics platforms from search and social to retail media and DSPs
- Optimisation is claimed at ad-group level, a finer grain than the campaign-level control most tools expose
- European vendor: Italian law, an EU-registered company, data storage declared in AWS Paris and the Italian Garante as supervisory authority
- Customer data ownership stated plainly in the terms, with a thirty-day export window after termination and a DPA available on request
Cons
- No public pricing whatsoever: no plans, no entry price and no currency are published anywhere on the site
- High eligibility threshold, since all three products are advertised from one million euros of annual media investment
- Neither a free plan nor a free trial is offered or mentioned
- Ads GPT is still an Early Adopters programme rather than a generally available module
- No security certification is published: no ISO 27001, no SOC 2 and no trust or security page, only measures described in prose
- No subprocessor list is published; it is supplied only on request to the data controller
- Every performance figure comes from anonymised case studies, with no named client and no published methodology
Pricing & Plans
AD cube publishes no prices. The pricing page states that there are no standard plans and invites prospects to configure a package by declaring their annual advertising budget and selecting the products they need, after which a tailored proposal is sent, promised within twenty-four hours. Neither a permanent free plan nor a free trial is mentioned anywhere on the site. The only figures displayed, namely under one million euros, one to four million, five to eight million and above eight million, describe the customer's own annual media spend rather than the cost of the tool; the three products are advertised as available from one million euros of annual investment, which is an eligibility threshold and not a price.
Data, GDPR & hosting
A consolidated view of how AD cube handles your data.
GDPR overview
GDPR implementation is concrete rather than declarative. AD cube S.r.l. is named as data controller with its address, tax code, email and telephone number. The privacy policy states it was written in compliance with Regulation (EU) 2016/679 and applicable Italian law, sets out a legal basis and a retention period for each purpose, and lists the rights of access, rectification, erasure, restriction under Article 18, portability under Article 20, withdrawal of consent and objection under Article 21, together with the right to complain to the Italian Garante under Article 77. Transfers outside the European Union are conditioned on the safeguards of Articles 45, 46 and 49. The terms appoint AD cube as processor, or provide for a separate DPA at the customer's request, and require breach notification without undue delay. No data protection officer is named.
Who owns the data?
The customer keeps all rights, title and interest in its data, and AD cube states plainly that it claims no ownership of it. What the customer grants instead is a limited, non-exclusive, royalty-free licence, valid only for the life of the contract, letting AD cube host, process and use that data solely as far as is necessary to deliver the service. Outputs such as reports, MMM models and forecasts also remain at the customer's disposal, while the software, algorithms and AI models stay AD cube's property. Data may reach IT, storage, tax and legal suppliers acting as processors or independent controllers; the vendor states it does not sell or transfer user data to third parties for AI modelling.
Reuse rights
Customers can reuse what comes out of the platform without asking permission. The terms define outputs as the insights, reports, MMM models, budget allocations, forecasts, recommendations and analyses generated from customer data, and state that both those outputs and the underlying data remain at the customer's full disposal, AD cube being allowed to use them only to provide the service. The licence granted in the other direction is narrower: access to the platform itself is limited, non-exclusive, non-transferable and non-sublicensable, and is restricted to the customer's internal business purposes for the contract period. Reselling or transferring access without written consent, building a competing product, reverse-engineering the platform and scraping it are all prohibited. On termination the customer has thirty days to request an export of its data.
Data retention & training
Hosting summary
The privacy policy states that data is stored using cloud computing on servers located inside the European Union, specifically AWS Paris in availability zone eu-west-3b. Its stated scope covers www.adcube.ai and its subdomains, which takes in the app.adcube.ai platform. Processing is described as taking place mainly in Italy and the wider European Union, although transfers outside the Union are possible; where they occur they must rest on the safeguards of Articles 45, 46 and 49 of the GDPR, meaning an adequacy decision, binding corporate rules or standard contractual clauses. Customers may request a copy of any data held abroad and details of where it is stored. On the security side, the terms describe encryption in transit, role-based access controls, environment segregation, security event logging and monitoring, and an incident management process, while outages at a third-party cloud provider are treated as force majeure. One distinction is worth keeping in mind: the marketing site itself resolves to an address in Frankfurt, Germany, on Hostinger infrastructure, which is separate from where platform data is stored. No hosting certification is claimed.
Things to keep in mind
Risks and trade-offs to weigh before adopting AD cube.
- Handing budget decisions to a model: the terms warn that outputs are probabilistic and that the customer alone is responsible for business decisions taken on the strength of them
- The service is supplied on an as is and as available basis, with no warranty of any economic result and none of uninterrupted operation
- Dependence risk: with auto-pilot reallocating spend daily, a team can gradually lose its own working grasp of the trade-offs being made on its behalf
- Headline results such as thirty per cent more ROI, thirty to sixty per cent creative efficiency and ten times ROAS come from anonymised case studies with no published methodology
- Audience Discovery works by scraping social pages and behavioural data, so compliance with the GDPR and with the source platforms' own terms deserves checking before use
- Legal documentation is inconsistent: the terms page carries two overlapping versions of the contract with two different Milan addresses, and the Italian-language cookie policy links to a third party's privacy policy
- The web application's backend at api.adcube.ai publicly exposes Swagger, ReDoc and a full OpenAPI description including administrative routes, which is not an advertised product feature
Setup & Integrations
Technical difficulty
Technically the setup is light. Connecting an advertising account is described as a single click, campaigns are then imported and an optimisation goal is set; Creative DNA advertises no complex setup and results within minutes. There is nothing to install, no browser extension and no SDK to integrate. The real barrier is commercial rather than technical, since access begins with a sales conversation, a tailored quote and a signed subscription form. The marketing mix model benefits from historical data, though the vendor claims it copes with any data quality.
Deployment
Integrations
Behind AD cube
Social
Resources
All the official URLs gathered for verification and reference.
Frequently asked questions
What exactly is AD cube?
Who is behind the product?
How much does AD cube cost?
Is there a free trial or a free plan?
Is there a minimum advertising budget?
Which advertising platforms can I connect?
Where is my data hosted?
Who owns the data I put into the platform?
What happens to my data when the contract ends?
Is there a mobile app or a public API?
Should you pick AD cube?
AD cube is a serious piece of advertising technology aimed at a narrow audience. Its distinguishing feature is coverage: rather than solving one slice of the problem, it strings together annual planning through a marketing mix model, daily budget reallocation at ad-group level, and creative analysis that tells a team which visual and copy elements actually move performance. The academic lineage is real and checkable, running back to the Politecnico di Milano and to co-founders who publish in the field, and the claim that the mix model tolerates messy data addresses a genuine obstacle for advertisers who do not have clean historical series.
The reservations are equally clear. Nothing about pricing is public, and the entry threshold advertised, one million euros of annual media investment, rules out most advertisers before any conversation begins. There is no free plan and no trial, one module is still in an Early Adopters programme, and no security certification is published, only security measures described in prose. Every performance figure on the site comes from anonymised case studies without a named client or a stated methodology, so the headline numbers should be treated as marketing claims rather than benchmarks.
For an advertiser already spending at that level, the European footing matters: an Italian company, Italian governing law, data storage declared in AWS Paris, plain statements of data ownership and a thirty-day export window. Anyone considering it should press for the specifics that the site withholds, namely what a package actually costs, which products are included, what the contractual service levels are, and how model outputs are validated before budget follows them.
- Choosing a selection results in a full page refresh.
- Opens in a new window.