<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>ProAnalytiqs — Insights</title>
    <link>https://proanalytiqs.com/insights/</link>
    <description>Business intelligence, data modelling, and the AI that stands on them.</description>
    <language>en-GB</language>
    <atom:link href="https://proanalytiqs.com/rss.xml" rel="self" type="application/rss+xml"/>
    <item>
      <title>First payment default, and the agent that watches it</title>
      <link>https://proanalytiqs.com/insights/kpi-first-payment-default/</link>
      <guid isPermaLink="true">https://proanalytiqs.com/insights/kpi-first-payment-default/</guid>
      <pubDate>Fri, 28 Aug 2026 09:00:00 GMT</pubDate>
      <description>FPD is the earliest honest signal that underwriting or acquisition changed. What it means commercially, how to define it, and what an agent does with it.</description>
      <category>KPI series</category><category>Risk</category>
    </item>
    <item>
      <title>The recovery funnel, and the number nobody wants to measure</title>
      <link>https://proanalytiqs.com/insights/kpi-recovery-funnel/</link>
      <guid isPermaLink="true">https://proanalytiqs.com/insights/kpi-recovery-funnel/</guid>
      <pubDate>Thu, 27 Aug 2026 09:00:00 GMT</pubDate>
      <description>Delinquent, right party contact, promise to pay, promise kept. What each stage is worth, and why self-cure decides your collections budget.</description>
      <category>KPI series</category><category>Collections</category>
    </item>
    <item>
      <title>How every analytics role changed in eighteen months</title>
      <link>https://proanalytiqs.com/insights/how-analytics-roles-changed/</link>
      <guid isPermaLink="true">https://proanalytiqs.com/insights/how-analytics-roles-changed/</guid>
      <pubDate>Wed, 26 Aug 2026 09:00:00 GMT</pubDate>
      <description>Not replaced — rewritten. What the analytics engineer, BI developer, analyst, data engineer and data scientist actually do now, and the three roles that did not exist before.</description>
      <category>Teams</category><category>AI</category>
    </item>
    <item>
      <title>Sales pacing, and why the monthly forecast review is three weeks too late</title>
      <link>https://proanalytiqs.com/insights/kpi-sales-pacing/</link>
      <guid isPermaLink="true">https://proanalytiqs.com/insights/kpi-sales-pacing/</guid>
      <pubDate>Tue, 25 Aug 2026 09:00:00 GMT</pubDate>
      <description>Issued against a pro-rata forecast is the fastest honest read on whether the commercial plan is real. How to build it so it does not lie in week one.</description>
      <category>KPI series</category><category>Sales</category>
    </item>
    <item>
      <title>Acquisition source quality, or why your cheapest channel is your most expensive</title>
      <link>https://proanalytiqs.com/insights/kpi-acquisition-source-quality/</link>
      <guid isPermaLink="true">https://proanalytiqs.com/insights/kpi-acquisition-source-quality/</guid>
      <pubDate>Fri, 21 Aug 2026 09:00:00 GMT</pubDate>
      <description>Cost per funded loan ranks channels one way. Contribution per cohort ranks them another. The gap is the largest unclaimed margin in the business.</description>
      <category>KPI series</category><category>Marketing</category>
    </item>
    <item>
      <title>Your semantic layer is your context layer</title>
      <link>https://proanalytiqs.com/insights/semantic-layer-is-your-context-layer/</link>
      <guid isPermaLink="true">https://proanalytiqs.com/insights/semantic-layer-is-your-context-layer/</guid>
      <pubDate>Wed, 19 Aug 2026 09:00:00 GMT</pubDate>
      <description>What stands between a useful AI assistant and a confident liar is not the model. It is whether your metrics are defined once, tested and described.</description>
      <category>Context</category><category>dbt</category>
    </item>
    <item>
      <title>Onboarding drop-off, and the identity check that costs more than a pricing decision</title>
      <link>https://proanalytiqs.com/insights/kpi-onboarding-dropoff/</link>
      <guid isPermaLink="true">https://proanalytiqs.com/insights/kpi-onboarding-dropoff/</guid>
      <pubDate>Tue, 18 Aug 2026 09:00:00 GMT</pubDate>
      <description>A verification step that starts timing out costs more in a week than most pricing decisions save in a quarter. How to find out the same morning.</description>
      <category>KPI series</category><category>Product</category>
    </item>
    <item>
      <title>Cash, yield and the reconciliation nobody wants to build</title>
      <link>https://proanalytiqs.com/insights/kpi-cash-and-yield/</link>
      <guid isPermaLink="true">https://proanalytiqs.com/insights/kpi-cash-and-yield/</guid>
      <pubDate>Fri, 14 Aug 2026 09:00:00 GMT</pubDate>
      <description>The gap between your risk reporting and the general ledger is always there. Reported monthly by component it is a control. Discovered by a funder it is a finding.</description>
      <category>KPI series</category><category>Finance</category>
    </item>
    <item>
      <title>The KPI framework we build for a consumer lender in the first 30 days</title>
      <link>https://proanalytiqs.com/insights/kpi-framework-for-a-consumer-lender/</link>
      <guid isPermaLink="true">https://proanalytiqs.com/insights/kpi-framework-for-a-consumer-lender/</guid>
      <pubDate>Wed, 12 Aug 2026 09:00:00 GMT</pubDate>
      <description>The metric set we put in first for a consumer lender, in the order we build it — and the four definitions that cause the most arguments.</description>
      <category>Lending</category><category>Business intelligence</category>
    </item>
    <item>
      <title>Text-to-SQL fails on your warehouse, and it is not the model's fault</title>
      <link>https://proanalytiqs.com/insights/text-to-sql-fails-on-your-warehouse/</link>
      <guid isPermaLink="true">https://proanalytiqs.com/insights/text-to-sql-fails-on-your-warehouse/</guid>
      <pubDate>Wed, 05 Aug 2026 09:00:00 GMT</pubDate>
      <description>Benchmarks are run on tidy schemas with clean names. Your warehouse has thirteen years of history in it. Here is where the accuracy actually goes, and what to do instead.</description>
      <category>Context</category><category>AI</category>
    </item>
    <item>
      <title>Putting AI in a credit decision without losing the argument later</title>
      <link>https://proanalytiqs.com/insights/ai-in-a-credit-decision/</link>
      <guid isPermaLink="true">https://proanalytiqs.com/insights/ai-in-a-credit-decision/</guid>
      <pubDate>Sat, 01 Aug 2026 09:00:00 GMT</pubDate>
      <description>Creditworthiness scoring is an Annex III high-risk use under the EU AI Act. What that means in engineering terms, not legal ones.</description>
      <category>Guardrails</category><category>Lending</category>
    </item>
    <item>
      <title>Why nobody opens your Power BI reports</title>
      <link>https://proanalytiqs.com/insights/why-nobody-opens-your-power-bi-reports/</link>
      <guid isPermaLink="true">https://proanalytiqs.com/insights/why-nobody-opens-your-power-bi-reports/</guid>
      <pubDate>Wed, 29 Jul 2026 09:00:00 GMT</pubDate>
      <description>Low dashboard adoption is almost never a training problem. It is usually one of five specific failures in the model underneath — and four of them are fixable in a fortnight.</description>
      <category>Business intelligence</category><category>Power BI</category>
    </item>
    <item>
      <title>Getting a Power BI semantic model ready for Copilot</title>
      <link>https://proanalytiqs.com/insights/power-bi-copilot-readiness/</link>
      <guid isPermaLink="true">https://proanalytiqs.com/insights/power-bi-copilot-readiness/</guid>
      <pubDate>Wed, 22 Jul 2026 09:00:00 GMT</pubDate>
      <description>Most Copilot disappointment is a model preparation problem wearing a Copilot costume. The checklist we work through before switching it on for a business.</description>
      <category>Power BI</category><category>Context</category>
    </item>
    <item>
      <title>Evals are the new unit tests</title>
      <link>https://proanalytiqs.com/insights/evals-are-the-new-unit-tests/</link>
      <guid isPermaLink="true">https://proanalytiqs.com/insights/evals-are-the-new-unit-tests/</guid>
      <pubDate>Wed, 15 Jul 2026 09:00:00 GMT</pubDate>
      <description>If you cannot say whether last week's prompt change made the system better or worse, you are not engineering. The smallest eval suite worth having.</description>
      <category>Guardrails</category><category>AI</category>
    </item>
    <item>
      <title>Credit risk reporting that survives the questions after it</title>
      <link>https://proanalytiqs.com/insights/credit-risk-reporting-that-survives-scrutiny/</link>
      <guid isPermaLink="true">https://proanalytiqs.com/insights/credit-risk-reporting-that-survives-scrutiny/</guid>
      <pubDate>Wed, 08 Jul 2026 09:00:00 GMT</pubDate>
      <description>A regulator, a funder and an auditor each ask a different question about the same number. What it takes to answer all three without a spreadsheet in the middle.</description>
      <category>Risk</category><category>Lending</category>
    </item>
    <item>
      <title>Databricks or Snowflake for a lender's data platform</title>
      <link>https://proanalytiqs.com/insights/databricks-or-snowflake-for-a-lender/</link>
      <guid isPermaLink="true">https://proanalytiqs.com/insights/databricks-or-snowflake-for-a-lender/</guid>
      <pubDate>Wed, 24 Jun 2026 09:00:00 GMT</pubDate>
      <description>Both will work. The choice usually comes down to four things that have nothing to do with the feature comparison tables, and one of them is who you can hire in your city.</description>
      <category>Data engineering</category><category>Platform</category>
    </item>
    <item>
      <title>Analytics pods versus staff augmentation — the difference is who does the thinking</title>
      <link>https://proanalytiqs.com/insights/analytics-pods-versus-staff-augmentation/</link>
      <guid isPermaLink="true">https://proanalytiqs.com/insights/analytics-pods-versus-staff-augmentation/</guid>
      <pubDate>Wed, 03 Jun 2026 09:00:00 GMT</pubDate>
      <description>Three contractors and a three-person pod cost about the same and produce different outcomes. The variable is where the integration work happens.</description>
      <category>Delivery</category><category>Teams</category>
    </item>
    <item>
      <title>The event taxonomy conversation nobody wants to have</title>
      <link>https://proanalytiqs.com/insights/event-taxonomy-conversation-nobody-wants/</link>
      <guid isPermaLink="true">https://proanalytiqs.com/insights/event-taxonomy-conversation-nobody-wants/</guid>
      <pubDate>Fri, 15 May 2026 09:00:00 GMT</pubDate>
      <description>Product analytics fails at the naming stage, months before anyone looks at a funnel. The two-hour conversation that prevents it, and the rules that last.</description>
      <category>Product analytics</category><category>Data engineering</category>
    </item>
  </channel>
</rss>