How Long Does Improvado Actually Take to Set Up?

How long does Improvado take to set up, and when do you actually see value from it? The connection step for a single platform is a five-minute OAuth flow, not a weeks-long project, and the self-serve tiers put you inside the product the same day you sign up. Total time to full value depends on how many platforms you connect and how each one authenticates, which this article breaks down with sourced numbers instead of a vague estimate.

Key Takeaways

• Connecting a single data source through Improvado is an OAuth flow that Improvado's own connector pages describe as completing in under five minutes for platforms like Google Ads.
• Improvado's connection documentation lays out three authentication paths, OAuth for platforms like Google Ads and Facebook, login credentials for others, and API tokens for the rest, and the path you get determines how fast that one connection goes.
• The Free Limited tier is $0 a month with immediate AI Agent and MCP access, and the MCP Only tier is $100 a month, so a first working connection does not require a sales cycle.
• ASUS reports Improvado saves the team about 90 hours a week that used to go into manually standardizing regional reports.
• AdCellerant cut the engineering time to add a new ad-platform integration from over six months to three, a 70 percent reduction in integration cost per platform.
• The AI-agent era changes what "set up" even means: instead of waiting for a dashboard build, an agent connected through Improvado's MCP layer can query data the moment a source is connected.

What "Setting Up Improvado" Actually Involves

The unit of work in any onboarding is a single data source connection, and Improvado's own documentation on connecting data is specific about what that involves. You open the Data Sources page, pick a platform, and authenticate. The credentials required depend on the platform: some providers, Google Ads and Facebook among them, use OAuth, and if you are already logged into the right account, the docs describe that step as happening "in a blink of an eye." Others require a login and password you already hold in that platform's own system. A smaller set, Criteo is the documentation's example, need an additional API token you retrieve from the source itself. None of these three paths involve engineering time on your side, and none of them involve Improvado building custom code to reach your account.

Once a source is connected, it is available to any BI tool or destination you point at it, and the individual connector pages carry this through into an actual number: authenticating a platform like Google Ads over OAuth is described as completing in under five minutes, after which your data starts normalizing against the same schema as every other connected source. That is the true unit of onboarding speed, one connection, one authentication step, minutes rather than a project.

What Actually Determines How Fast You Get There

If a single connection is a five-minute task, then total time to value is mostly a function of three things, and none of them are secret.

The first is how many sources you are connecting. A team running ads on three platforms and a CRM is doing four connection steps. A multinational advertiser running dozens of regional accounts across a dozen platforms, the kind of setup ASUS describes, is doing a larger number of the same repeatable step, not a fundamentally different kind of work.

The second is which authentication path each source uses. OAuth sources are the fastest because the credential exchange happens between the platform and Improvado directly. Login-based and token-based sources add a step, retrieving the token or having the password on hand, but they are still a single connection, not a build.

The third is which pricing tier you start on, because that determines whether you are moving at your own pace or coordinating a sales and implementation process. On the Free Limited tier, at $0 a month, live API requests span all sources and the AI Agent and MCP layer is included from the first login, with a cap of 50 MCP actions a week and one workspace. The MCP Only tier, at $100 a month, raises that to 300 MCP actions a week, daily data sync, and two million rows a year, still entirely self-serve. Advanced and Enterprise plans are quoted to the customer's scale rather than published as a flat number, because at that volume, hundreds of sources and custom governance rules, the work genuinely does vary by scope. Enterprise-tier deployments elsewhere on the site are described starting at $30,000-plus annually, about $2,500 a month, priced against data volume and source count, the same two variables that determine connection time in the first place.

What This Looks Like for Real Customers

Two published case studies put numbers on what changes once the connections exist, and both are about ongoing time saved rather than a one-time setup event, which is the more honest way to measure it.

ASUS had to consolidate marketing reporting across Western Europe, Eastern Europe, South America, North America, APAC, and Taiwan, with each region producing reports at different speeds and in different formats. Before Improvado, standardizing that into one comparable view took days per cycle and required pulling in a technical team to do it. After connecting their sources, the company's own account of the result is direct: "we don't have to involve our technical team in the reporting part at all. Improvado saves about 90 hours per week and allows us to focus on data analysis rather than routine data aggregation, normalization, and formatting."

AdCellerant is a different kind of case, because it measures the time and cost of building a data integration, the same work that determines how fast any platform onboards a new data source in the first place. Before switching to Improvado's embedded integrations, AdCellerant's engineering team spent over six months and roughly $120,000 building each new ad-platform connector in-house. With Improvado's connectors already built and its embedded white-label option in place, that dropped to about three months and $40,000 per platform, a 70 percent cost reduction that the team attributes directly to not rebuilding what a connector platform already provides.

How the AI-Agent Era Changes the Starting Line

The older model of onboarding treated a connected data source as a step toward eventually building a dashboard, which meant the useful outcome, an answer to a business question, arrived after both the connection work and the reporting work were done. MCP access changes what happens between those two steps.

Improvado's MCP gateway gives an AI agent, whether that is Claude, ChatGPT, or another agent a team already uses, governed access to every connected source through one endpoint rather than a key per platform. Practically, that means the moment a source is connected and normalized, an agent can query it directly, without a human building a chart first. The MCP Only tier exists specifically for teams that want this path without adopting the full BI product: connections, a knowledge graph, and modeling for $100 a month, using whatever AI agent setup the team already has.

The knowledge graph layer is what makes that querying reliable rather than a guess. Every platform names its own metrics and entities differently, so an agent asking a question across several connected sources needs one consistent set of definitions to query against, which is what the knowledge graph functions as: shared memory an agent reads from and writes to, instead of each agent holding its own private interpretation of what a metric means. Combined with the standard 14-day free trial with no credit card required, a team can test this entire path, connect a source, and have an agent query it, before spending anything.

The Honest Answer

There is no single published number for "days to full Improvado deployment" because it depends on how many sources a team is connecting and which tier they start on, and any platform that claims otherwise is describing a demo, not a deployment. What is sourced and verifiable is this: the unit of onboarding work, connecting one data source, is a five-minute OAuth flow for the platforms that support it, not a professional-services engagement. The self-serve tiers put that first connection in front of you the same day you sign up, at $0 or $100 a month, with no sales process required to start. And the ongoing payoff once sources are connected is measured in real customer accounts, not projections, ASUS reclaiming close to 90 hours a week of manual reporting, AdCellerant cutting integration costs by 70 percent per platform.

See how fast your own data sources connect, start with the free tier or talk to an Improvado expert.