<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0">
    <channel>
        <title>Signals | Tricky Wombat</title>
        <link>https://www.trickywombat.ai/signals</link>
        <description>Research-backed analysis on AI search, chatbots, and the decisions that matter for mid-market companies.</description>
        <lastBuildDate>Fri, 03 Apr 2026 12:38:32 GMT</lastBuildDate>
        <docs>https://validator.w3.org/feed/docs/rss2.html</docs>
        <generator>https://github.com/jpmonette/feed</generator>
        <language>en</language>
        <image>
            <title>Signals | Tricky Wombat</title>
            <url>https://www.trickywombat.ai/images/tricky-wombat-1200x630.png</url>
            <link>https://www.trickywombat.ai/signals</link>
        </image>
        <copyright>All rights reserved 2026, Tricky Wombat</copyright>
        <item>
            <title><![CDATA[How people use AI matters more than which AI they use]]></title>
            <link>https://www.trickywombat.ai/signals/why-design-matters-for-ai</link>
            <guid isPermaLink="false">https://www.trickywombat.ai/signals/why-design-matters-for-ai</guid>
            <pubDate>Fri, 03 Apr 2026 03:03:14 GMT</pubDate>
            <description><![CDATA[Organizations spending identical amounts on identical AI models see returns ranging from 2% to 300%. The variable is how the product is designed, not which model powers it.]]></description>
            <author>Tricky Wombat Research Team</author>
        </item>
        <item>
            <title><![CDATA[What exactly IS prompt engineering?]]></title>
            <link>https://www.trickywombat.ai/signals/what-is-prompt-engineering</link>
            <guid isPermaLink="false">https://www.trickywombat.ai/signals/what-is-prompt-engineering</guid>
            <pubDate>Mon, 30 Mar 2026 23:32:42 GMT</pubDate>
            <description><![CDATA[Prompt engineering is where AI results start, not where they peak]]></description>
            <author>Tricky Wombat Research Team</author>
        </item>
        <item>
            <title><![CDATA[How to roll out AI implementations]]></title>
            <link>https://www.trickywombat.ai/signals/technology-roll-out</link>
            <guid isPermaLink="false">https://www.trickywombat.ai/signals/technology-roll-out</guid>
            <pubDate>Mon, 30 Mar 2026 22:48:47 GMT</pubDate>
            <description><![CDATA[Are there techniques that high-performing organizations use to roll out their AI implementations? Lets explore finding a special recipe that all organizations can strive for.]]></description>
            <author>Tricky Wombat Research Team</author>
        </item>
        <item>
            <title><![CDATA[The real benefits of adding an AI chatbot to your website.]]></title>
            <link>https://www.trickywombat.ai/signals/benefits-of-ai-chatbot</link>
            <guid isPermaLink="false">https://www.trickywombat.ai/signals/benefits-of-ai-chatbot</guid>
            <pubDate>Tue, 17 Mar 2026 06:10:26 GMT</pubDate>
            <description><![CDATA[Companies with strong knowledge bases achieve 75% AI resolution rates and 369% ROI. The 39% that fail almost always have an infrastructure problem, not a model problem.]]></description>
            <author>Tricky Wombat Research Team</author>
        </item>
        <item>
            <title><![CDATA[Prompt engineering is only the first step to create great AI]]></title>
            <link>https://www.trickywombat.ai/signals/prompt-engineering</link>
            <guid isPermaLink="false">https://www.trickywombat.ai/signals/prompt-engineering</guid>
            <pubDate>Sun, 29 Mar 2026 18:50:44 GMT</pubDate>
            <description><![CDATA[The first five hours of prompt work produce a 35% accuracy gain. The next 60 add only 6%. The best organizations treat prompt engineering as only the first step.]]></description>
            <author>Tricky Wombat Research Team</author>
        </item>
        <item>
            <title><![CDATA[Context engineering]]></title>
            <link>https://www.trickywombat.ai/signals/context-engineering</link>
            <guid isPermaLink="false">https://www.trickywombat.ai/signals/context-engineering</guid>
            <pubDate>Sat, 28 Mar 2026 07:09:32 GMT</pubDate>
            <description><![CDATA[A well engineered pipeline matters. The gap between relevant response and poor answers is not a model problem. A smaller AI with well-engineered context outperforms a much bigger model without it. Context engineering, the discipline of designing what an AI sees before it responds, explains the gap.]]></description>
            <author>Tricky Wombat Research Team</author>
        </item>
        <item>
            <title><![CDATA[Your AI support bot isn't stupid]]></title>
            <link>https://www.trickywombat.ai/signals/bot-isnt-stupid</link>
            <guid isPermaLink="false">https://www.trickywombat.ai/signals/bot-isnt-stupid</guid>
            <pubDate>Fri, 20 Mar 2026 06:02:05 GMT</pubDate>
            <description><![CDATA[Peer-reviewed research shows the same LLM swings from 50% to 87% accuracy based solely on retrieval pipeline design. Yet 95% of enterprise AI pilots deliver zero P&L impact. The model was never the bottleneck.]]></description>
            <author>Tricky Wombat Research Team</author>
        </item>
        <item>
            <title><![CDATA[What is the ROI for AI customer service?]]></title>
            <link>https://www.trickywombat.ai/signals/what-is-the-roi-for-ai-customer-service</link>
            <guid isPermaLink="false">https://www.trickywombat.ai/signals/what-is-the-roi-for-ai-customer-service</guid>
            <pubDate>Mon, 09 Mar 2026 19:24:02 GMT</pubDate>
            <description><![CDATA[Klarna's chatbot projected $40M in profit improvement. That didn't materialize, so they started rehiring human agents within a year.]]></description>
            <author>Tricky Wombat Research Team</author>
        </item>
        <item>
            <title><![CDATA[The Real AI Chatbot Implementation Timeline]]></title>
            <link>https://www.trickywombat.ai/signals/chat-implementation-timeline</link>
            <guid isPermaLink="false">https://www.trickywombat.ai/signals/chat-implementation-timeline</guid>
            <pubDate>Fri, 20 Mar 2026 06:02:10 GMT</pubDate>
            <description><![CDATA[Gartner found 85% of CS leaders will pilot AI chatbots in 2025 but only 5% have deployed one. The bottleneck isn't the platform, it's the knowledge infrastructure underneath it.]]></description>
            <author>Tricky Wombat Research Team</author>
        </item>
        <item>
            <title><![CDATA[Good knowledge = good engineering]]></title>
            <link>https://www.trickywombat.ai/signals/good-engineering-knowledge-pipeline</link>
            <guid isPermaLink="false">https://www.trickywombat.ai/signals/good-engineering-knowledge-pipeline</guid>
            <pubDate>Mon, 09 Mar 2026 19:24:00 GMT</pubDate>
            <description><![CDATA[Most AI pilots fail not because the models are wrong but because engineering data is unclassified, ungoverned, and full of gaps. A problem that naive AI accelerates, not solves.]]></description>
            <author>Tricky Wombat Research Team</author>
        </item>
        <item>
            <title><![CDATA[AI in legal practice is broken]]></title>
            <link>https://www.trickywombat.ai/signals/ai-in-legal-practice-is-broken</link>
            <guid isPermaLink="false">https://www.trickywombat.ai/signals/ai-in-legal-practice-is-broken</guid>
            <pubDate>Mon, 09 Mar 2026 19:23:58 GMT</pubDate>
            <description><![CDATA[Legal AI tools hallucinate 17-33% of the time on verified legal databases. 486 documented fake citations are in court filings. The fix is your data pipeline, not your model.]]></description>
            <author>Tricky Wombat Research Team</author>
        </item>
        <item>
            <title><![CDATA[AI Slop is the New Spam]]></title>
            <link>https://www.trickywombat.ai/signals/ai-slop-is-new-spam</link>
            <guid isPermaLink="false">https://www.trickywombat.ai/signals/ai-slop-is-new-spam</guid>
            <pubDate>Mon, 09 Mar 2026 19:23:56 GMT</pubDate>
            <description><![CDATA[AI responses without the right information return the same things as SPAM. Generic, vague and formulaic. It is called SLOP for a reason. Learn how SLOP happens and how to fix it.]]></description>
            <author>Tricky Wombat Research Team</author>
        </item>
    </channel>
</rss>