What SIGNLD does not do

SIGNLD is not a data warehouse, does not write back to source systems, does not run custom statistical modelling, cannot answer about systems it is not connected to, is not built for very small companies with few systems, and is not a dashboard design tool.

By Christopher Rafter · · 8 min read · Product explained
What SIGNLD does not do

In this article

It is not a data warehouse or a system of record

SIGNLD does not store your operational data as a system of record, and it is not a replacement for a data warehouse. It reads from your existing systems, resolves entities across them in the Knowledge Graph, and answers questions. It does not become the place where your financial, sales, or operational data lives permanently.

For the wider context, see our overview of what SIGNLD is.

If your company needs a central warehouse for reasons beyond answering leadership questions, such as feeding a separate business intelligence tool or supporting a data science team's own pipelines, SIGNLD does not replace that infrastructure. It works alongside it or instead of it for the specific job of answering questions, not as a general-purpose data store.

A company that already runs a warehouse can still connect it to SIGNLD as one more source, and SIGNLD will read from it the same way it reads from a CRM or an accounting system. But SIGNLD does not take over the warehouse's job of storing raw history for every downstream use case a company might have.

It does not write back to source systems

Every connection SIGNLD makes is read-only. It cannot create, edit, or delete records in QuickBooks, a CRM, a support platform, or any other connected system. If a Decision Brief recommends an action, a person still has to go into the source system and make the change.

This means SIGNLD cannot automate a workflow that requires updating a record, sending an email, or triggering a downstream process. It surfaces the finding and the recommended action. It does not execute the action.

This is a deliberate boundary, not a missing feature waiting to be built. A read-only connection removes an entire category of risk: there is no scenario where a bad question, a bug, or a bad actor with account access could change a customer record, a financial figure, or a deal stage in a system the company depends on.

It does not run custom statistical or ML modelling

SIGNLD does not function as an analyst's notebook. It does not run custom regressions, build forecasting models, or let a data scientist write arbitrary Python or R against the connected data. The Decision Brief format is fixed: a finding, evidence, a confidence score, and a recommended action, generated in response to a specific question.

If your team needs to build and iterate on a custom model, such as a churn prediction model tuned with specific features and a specific algorithm, that work happens outside SIGNLD in a dedicated modelling environment. SIGNLD is not that environment.

SIGNLD can still surface the inputs a data science team would want to look at, since a Decision Brief traces evidence back to specific source records across systems. But building, training, and validating a custom model on top of that data is a separate workflow that happens in a separate tool.

It cannot answer about a system that is not connected

SIGNLD only knows what it can read. If a system is not connected, no question that depends on it will produce a real answer. There is no way to ask about data sitting in a tool that has not been set up as a source.

This sounds obvious, but it matters in practice: a question that spans finance, sales, and operations will only be as complete as the systems that are connected. If the operations system is missing, a Decision Brief about operational risk will be incomplete or unavailable, not filled in from guesswork.

SIGNLD will not fabricate a plausible-sounding answer to cover a gap. If a question depends on a system that is not connected, the honest response is that the question cannot be fully answered yet, along with what connecting that system would take. Read /connectors to check whether a specific system is supported before assuming a question can be answered.

It is not built for companies under 10 employees with two systems

SIGNLD is built to resolve entities and answer questions across many connected systems, typically ten or more as a company scales. A company with two systems and under ten employees can technically connect SIGNLD, but the value of cross-system resolution and Decision Briefs is smaller when there is little to cross. In that case, checking the two systems directly is often just as fast.

The product is aimed at the point where a company has enough systems, enough people asking questions, and enough data spread across sources that manually checking each one becomes the bottleneck. That threshold varies by company, but it is rarely a company running its entire operation out of two connected tools and a shared inbox.

It is not a pixel-level dashboard design tool

SIGNLD does not let you design a custom dashboard layout, pick chart types, or adjust colors and fonts on a visualization. It does not compete with dedicated business intelligence tools built around visual design and dashboard building.

If your team's requirement is a polished, customizable dashboard for an executive presentation or a customer-facing report, that is not what SIGNLD produces. A Decision Brief is a written finding with linked evidence, not a chart.

Teams that need both still typically use SIGNLD to find and confirm the answer, then build the visual version separately in a dedicated business intelligence or presentation tool once the underlying finding is settled.

What SIGNLD is actually for

SIGNLD is for a leadership team that wants a direct answer to a specific business question, backed by evidence from the systems the company already runs, without building a warehouse or waiting on a report. Read /why-us for the reasoning behind that focus and /how-it-works for the mechanics of how a question becomes a Decision Brief.

It connects to systems read-only, resolves the same customers, deals, and records across them, and returns a finding with a confidence score and a recommended action. That is the whole scope. Everything listed above is outside that scope on purpose.

The boundaries described in this post are not gaps to be filled in a future release. They are the edges of a deliberately narrow job: read what already exists across your systems, resolve it into a coherent picture, and answer the question a leader actually asked, with the evidence to check the answer.

Related reading in this series: What systems does SIGNLD connect to and What you get on the SIGNLD free plan.

Try SIGNLD free

See what SIGNLD can answer for your systems, and what it cannot. Try SIGNLD free.

Key takeaways

  • SIGNLD does not store your operational data as a system of record, and it is not a replacement for a data warehouse.
  • SIGNLD does not function as an analyst's notebook.
  • SIGNLD does not let you design a custom dashboard layout, pick chart types, or adjust colors and fonts on a visualization.
  • SIGNLD is for a leadership team that wants a direct answer to a specific business question, backed by evidence from the systems the company already runs, without building a warehouse or waiting on a report.
  • See what SIGNLD can answer for your systems, and what it cannot.