SIGNLD week one: what to expect

Day one, you connect one system and get a first answer within minutes. By day five, most teams have three to five systems connected, a working Knowledge Graph, and several Decision Briefs answered. Week two is when cross-system questions start.

Day one: which system to connect first

Most teams connect their financial system or their CRM first, because those two systems answer the most immediate leadership questions: revenue, pipeline, and cash. Read /connectors for the full list of 800-plus supported systems, but the choice usually comes down to whichever system holds the metric someone in the company is already asking about weekly.

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

Connecting the first system takes about 15 minutes. You authenticate a read-only connection, SIGNLD pulls the schema, and it starts building the first layer of the Knowledge Graph from that one source. Nothing is written back to the system. SIGNLD only reads.

At this stage the Knowledge Graph is a single-source map. It holds the tables, fields, and records from that one system, organized so a question can be matched to the right data. You can already ask direct questions about that system, but nothing that requires combining it with another source yet.

A typical day-one question is narrow on purpose: total revenue this month, the ten largest open invoices, or the number of deals currently sitting in a specific pipeline stage. These questions do not need entity resolution across systems, so they return quickly and give the team a sense of how a Decision Brief reads before anything more complex is asked. Nobody on the team needs to write a query or open the source system directly to get the answer.

Day two: your first Decision Brief

By day two, most teams ask their first real question and get a Decision Brief back. A Decision Brief has four parts: the finding, the evidence linked back to the specific source records, a confidence score, and a recommended action.

A typical first brief answers something narrow and concrete, such as which customer accounts have overdue invoices past 60 days, or which deals in the pipeline have had no activity in three weeks. The evidence section links directly to the invoice or deal records the finding is based on, so anyone reviewing the brief can check the underlying rows themselves rather than trusting a summary.

This is also the point where most teams learn how /how-it-works actually operates: SIGNLD does not pre-build dashboards. It answers the specific question you ask, when you ask it.

The first brief is also useful as a check on the connection itself. If the finding references the wrong date range or misses obvious records, that usually means the source system needs a permission adjustment or a field mapping check, not that the question was wrong. Most teams resolve this within the same day, since the read-only connection can be reviewed and reconnected without any risk to the source system.

Day three: adding a second system

Day three is usually when the second system goes in, often the CRM if finance went first, or the finance system if CRM went first. The second connection also takes about 15 minutes, and it does not require reconfiguring the first one.

Once two systems are connected, the Knowledge Graph starts doing more than storing two separate maps. It begins identifying where records in one system likely correspond to records in the other, such as a company name in the CRM matching a customer name in the accounting system. This resolution work is what eventually makes cross-system questions possible, but on day three it is still forming.

Questions you can ask on day three are still mostly single-system, plus a few basic joins where the entity names already match cleanly. A common day-three question is which accounts have both an open deal and an active invoice, which works reliably once both systems use the same customer name convention, but is not yet reliable for accounts with mismatched names or abbreviations.

Day four: the Knowledge Graph starts resolving entities

By day four, with two systems connected for a full day, the entity resolution layer has usually caught the obvious matches and started flagging the harder ones, like a customer listed as "Meridian Group" in the CRM and "Meridian Group LLC" in the accounting platform. SIGNLD treats these as the same underlying entity once resolved, which is what lets a Decision Brief pull evidence from both systems for a single customer.

This is the point where you can start asking questions that genuinely span two systems: which customers with open support tickets also have overdue invoices, or which deals closed last month have not yet been invoiced. The answers may still be incomplete if resolution has not caught every naming variant, but the pattern is working.

Confidence scores on these early cross-system briefs tend to be a little more conservative than they will be a week later, simply because the Knowledge Graph has had less time to confirm matches. That is expected and not a sign of a broken connection. It improves as more records pass through and more questions are asked.

Day five: a third or fourth system and cross-system questions

Many teams add a third system by day five, often a support platform, a project management tool, or a second data source like a spreadsheet tracking something the core systems do not capture. Spreadsheets are treated as first-class sources, not an afterthought, so a manually maintained cost tracker or a headcount plan works the same way a CRM connection does.

With three to five systems connected, the Knowledge Graph is wide enough that most day-to-day leadership questions can be answered without waiting on a report. A Decision Brief at this stage might trace a churn risk across CRM activity, support ticket volume, and invoice history for the same account, all in one answer, with evidence links into each source system.

This is usually the first day a non-technical leader, such as a CFO or head of operations, asks a question directly without going through whoever set up the connections. Since a Topic can be built around a specific recurring question, such as weekly cash position or open risk accounts, the team stops treating SIGNLD as a new tool to learn and starts treating it as a place to check before a meeting.

What week two looks like

Week two is usually when the questions get more specific and less exploratory. Instead of "what does this Decision Brief look like," the questions become "which accounts are at risk this quarter" or "which reps are behind on pipeline coverage." Entity resolution keeps improving as more records pass through the Knowledge Graph, and teams typically add one or two more systems as new questions surface gaps.

By the end of week two, most accounts have moved from testing SIGNLD on a narrow question to running it against the questions leadership actually asks in a weekly meeting. There is no requirement to build a data warehouse or migrate data anywhere first, since SIGNLD reads systems as they exist today.

Related reading in this series: What SIGNLD does not do and What systems does SIGNLD connect to.

Try SIGNLD free

Connect your first system and see your first Decision Brief today. Try SIGNLD free.

Key takeaways

Most teams connect their financial system or their CRM first, because those two systems answer the most immediate leadership questions: revenue, pipeline, and cash.
Day three is usually when the second system goes in, often the CRM if finance went first, or the finance system if CRM went first.
By day four, with two systems connected for a full day, the entity resolution layer has usually caught the obvious matches and started flagging the harder ones, like a customer listed as \"Meridian Group\" in the CRM and \"Meridian Group LLC\" in the accounting platform.
Week two is usually when the questions get more specific and less exploratory.
Connect your first system and see your first Decision Brief today.

In this article