LinkedIn thought leader ad examples

Der Kunde kündigt. Die KI gratuliert zum Jubiläum. Die persönlichste Nachricht der Welt kann im falschen Moment alles verschlimmern. Stell dir einen Kunden beim Telekommunikationsanbieter vor. Sein Anschluss fällt wiederholt aus. Er hat die Störung mehrfach gemeldet. Jetzt wartet er auf eine Rückmeldung. Dann kommt eine Nachricht. Mit seinem Namen, seinem Tarif und einem Treueangebot zum Vertragsjubiläum. Nur zum kaputten Anschluss steht nichts darin. Das Marketing freut sich über die sauber ausgespielte Kampagne. Der Kunde denkt sich: Für Werbung reicht es offenbar. Für eine Antwort nicht. Er kündigt. Seine nächste passende Ansprache wäre ein Service-Update gewesen. Mit einem verlässlichen Bearbeitungsstand und einem konkreten nächsten Schritt. Genau daran musste ich bei der angekündigten Braze-Übernahme durch Salesforce denken. Der vereinbarte Unternehmenswert liegt bei rund 5,66 Milliarden US-Dollar. Der Abschluss wird im Salesforce-Geschäftsjahr 2028 erwartet. Aktionärszustimmung und regulatorische Freigaben stehen noch aus. Die gemeinsame Lösung ist eine Zukunftsperspektive. Das Vorhaben soll Kundendaten und personalisierte Ansprache enger verbinden. Weniger Übergaben und schnellere Reaktionen sind ein sinnvolles Ziel. Ob daraus bessere Kundenbeziehungen werden, entscheidet sich zwischen dem eingehenden Signal und der nächsten Aktion. Für die Szene oben braucht es drei Dinge. Der offene Servicefall muss rechtzeitig bei der Versandentscheidung ankommen. Mit eindeutiger Kundenzuordnung und aktuellem Status. Ein Datenabgleich in der Nacht hilft wenig, wenn die Werbung am Nachmittag rausgeht. Marketing und Service brauchen eine gemeinsame Vorrangregel. Bei einer wiederholten, ungelösten Störung pausiert das Treueangebot. Der Service übernimmt die Rückmeldung. Auch bei fehlenden oder veralteten Daten muss feststehen, was passieren darf. Die Entscheidung muss nachvollziehbar bleiben. Es muss erkennbar sein, welches Signal welche Nachricht ausgelöst oder verhindert hat. Ebenso muss feststehen, wer die Regel ändern darf und wann Werbung wieder freigegeben wird. Aus meiner Sicht gehört genau das zur Kundenbindung: verlässliche Daten, klare Zuständigkeiten und Regeln, die das Unternehmen selbst erklären und verändern kann. KI kann bei der Auswahl helfen. Den Vorrang des Service muss das Unternehmen festlegen. Manchmal ist die beste Marketingaktion eine zurückgehaltene Werbenachricht. Wenn du eine solche Kundenstrecke prüfen möchtest, schreib mir. Wir schauen gemeinsam auf Servicesignal, Vorrangregel und Versandentscheidung.

𝗦𝗠𝗣 𝗿𝗲𝗮𝗰𝗵𝗲𝘀 𝗶𝘁𝘀 𝗵𝗶𝗴𝗵𝗲𝘀𝘁 𝗹𝗲𝘃𝗲𝗹 𝘀𝗶𝗻𝗰𝗲 𝟮𝟬𝟮𝟮 𝗮𝘀 𝗘𝘂𝗿𝗼𝗽𝗲𝗮𝗻 𝗽𝗼𝘄𝗱𝗲𝗿 𝗺𝗮𝗿𝗸𝗲𝘁𝘀 𝘀𝘁𝗿𝗲𝗻𝗴𝘁𝗵𝗲𝗻 The latest DCA Market Intelligence benchmarks show strong momentum in the European powder market. Food-grade SMP recorded the largest weekly increase among the tracked powder commodities and reached its highest level since 2022. Compared with last year, SMP benchmarks are now 78.7% to 81.6% higher. Sweet whey powder also moved higher, with Food SWP reaching the highest level ever recorded for the DCA benchmark. At the same time, WMP and lactose moved lower, while market contributors raised concerns about how higher European prices could affect competitiveness in international markets. What is driving the current strength in SMP, and what could it mean for the European powder market? Read the full DCA Market Intelligence analysis: https://lnkd.in/eTQqEZAg #DairyMarkets #SMP #MilkPowder #WheyPowder #DairyTrade #DairyBenchmarks #MarketIntelligence #DCAMarketIntelligence

𝗘𝗨 𝗱𝗮𝗶𝗿𝘆 𝗲𝘅𝗽𝗼𝗿𝘁𝘀 𝗮𝗿𝗲 𝗴𝗿𝗼𝘄𝗶𝗻𝗴 𝗳𝗮𝘀𝘁, 𝗯𝘂𝘁 𝗻𝗼𝘁 𝗳𝗮𝘀𝘁 𝗲𝗻𝗼𝘂𝗴𝗵 EU dairy exports increased by 11.88% year-on-year in August. Year-to-date, export volumes are now 7.18% above last year. But milk production has grown even faster. Imports are also falling, with August marking the seventh consecutive monthly decline. Even so, the additional milk has not been fully absorbed. As a result, estimated theoretical dairy stocks were 38.75% higher than at the start of the year. Read the full DCA Market Intelligence analysis: https://lnkd.in/eSsTVhQu #DairyMarkets #DairyTrade #DairyExports #MilkProduction #DairyIndustry #MarketIntelligence #DCAMarketIntelligence

If FINAL_v7 is your source of truth, Excel is running your metadata process. I've seen plenty of finance teams where one spreadsheet quietly became the thing everyone checked before making a master data change. One person updates ERP. Another updates EPM. The warehouse catches up later. At that point Excel isn't just holding metadata. It's deciding what's true across systems. And once a spreadsheet is making that call, the fix isn't a better spreadsheet. The change needs one governed path, from request to audit. That's what EPMware does for financial master data. Excel is useful for a lot of things. Deciding which version is right shouldn't be one of them.

Omni is a launch partner in the SpaceX Super Intelligence Partner Network 🤝 We love building in Cursor, and so do many of our customers. Omni's Cursor plugin and MCP server bring governed data into those workflows, so you can query, define new metrics, and build data apps on the shared definitions your team trusts without leaving your preferred workspace 🛠️ Here's what that looks like end to end. Peter on our Solutions Engineering team built a demo customer portal for our sales team, starting from one paragraph in Cursor and ending with embedded dashboards and AI chat 👉 https://lnkd.in/g3u9RfAH Thanks to the SpaceX team for the partnership, and check out this blog from Peter for a deeper dive on Omni + Cursor.

Have you ever sat through same boring eCommerce demo? I know I have. I've given it a billion times and was ready for something new. But how was I going to change that same old demo? I would have to build a fake demo dataset, transform that data in dbt Labs, model and create content in Omni, and build a fully functional embedded web application. My skill set only covers one of those four. This is where Cursor came in. Using Cursor I got to moonlight as a data engineer, analytics engineer and a software engineer in the same IDE. Check out my blog below walking you through how I used Cursor, dbt and Omni to build a fully functional AI powered data application for a fictional SaaS company, Data Blobby, that our SE teams could demo to prospects. Omni is excited to be partnering with SpaceX Super Intelligence in their Partner Network launch! See how he built it 👉 https://lnkd.in/g-4zSkf7

It's a pleasure to welcome Alex Merced from Dremio as an OSA Con 2026 speaker on November 2 in San Francisco. Alex will be introducing the "Open Agentic Lakehouse." https://lnkd.in/g-359vDS. It's in-person! Alex is the most eloquent speaker on data lake technologies I know. His talks offer an excellent overview of data lake architecture and the open source projects that make up the foundation. This year's presentation describes emerging open protocols to integrate AI agents with data lakes. It's must viewing for anyone building modern analytic systems. #opensource #agentic #lakehouse #apache #iceberg #ossie #polaris

A studio apologizes for a broken patch by sending players premium currency. Two weeks later, the agent flags runaway inflation. The agent wasn’t bad at SQL. The rule “never count compensation grants as a faucet” just lived in one person’s head. That’s what Edition 4 of Level Up is about: the rules your team applies by reflex and never wrote down, and what happens to them when an agent starts answering the questions. Inside: -> How to let corrections turn into rules, instead of trying to document everything up front -> Why every change to your metrics should go through review like a pull request -> How to use the agent’s own usage to find what’s missing in your context layer -> 8 questions to tell if your context is governed or just assembled

80% of our positive replies last month had one thing in common: good timing And honestly, it had nothing to do with the copy (all of our campaigns have proven copy framework and good offer that more or less all perform the same) It was simply the timing ↓ → they started a new job → they got a new title → they posted about the exact problem you solve → they commented on someone else's post about it That's pretty much the 4 moments we watch for. You can spot most of them yourself by scrolling LinkedIn….. it works, it just takes forever (or use a tool like Findymail Signals). I personally would be careful with job changes though (great on your dream accounts, I definitely wouldn't use them for volume). Timing beats messaging, every time. Well. Almost because (the offer still has to be good).

The new GRAX is here. Welcome to a universe where full ownership of your operational history is assumed rather than "extra". When customers own all versions of their historical data in their own private cloud, they unlock new possibilities for how data is governed, analyzed, integrated, and leveraged with AI. Insights that took weeks to uncover are now available in seconds. Business users who did not have access to complex querying capabilities now audit legacy data sets with precision. Want to surface CRM insights in Snowflake? We've got you covered! Need to unify data sets from multiple Salesforce instances? Check! Need data access without compromising security? You know we do that too! I could not be more proud to be part of this team...let's launch! Clinton Ford, Chris Cimaglia, Jason Wilburn, Joe Gaska, Ben Melton, Rich Bloyd, Diane Deng, Stuart Newton, Alex Blanes, Diego Bellalta, Ascendx, Ortoo, Documill, Eposly. https://www.grax.com/

Mining Ops - real-time tracking of trucks using #IconMap with #RealtimeIntelligence in #MicrosoftFabric. In earlier posts I showed tracking aircraft via ADS-B, then trains in Spain via GTFS. This example also uses our vehicle tracker map layer, but this time showing trucks as they travel up and down a mine in Chile. Using satellite imagery from Mapbox and Icon Map's built in digital elevation model (DEM) we can show the realtime locations within the mine highlighting any alerts from sensor data. Selecting on a truck can reveal additional information in a popup panel. We also support the ability to provide your own elevation models, 3D vehicle models and drone imagery for the most up to date and realistic digital twin. For more information sign up for the preview: https://lnkd.in/ecpnG2nU #IconMapForFabric #KQL #EventHouse #3DMaps

Using #QGIS to prep your map data for #MicrosoftFabric? With #IconMapForFabric, our QGIS plugin enables you to upload your map layers directly into #OneLake, either as a new Icon Map map, layers in an existing map or as new items in an organizational geospatial catalog. Want to try it out, sign up for the preview: https://lnkd.in/ecpnG2nU

Most QA teams don’t actually have a testing problem. They have a test data problem that cuts their sprint velocity in half. Ask an average QA Lead where most of their sprint time goes, and the answer is rarely "writing automated test scripts." It’s almost always: · Waiting on the DBA for a fresh database refresh. · Searching for hours to find that one specific customer edge case in a massive, polluted database. · Manually faking data in Excel or Notepad because crucial test scenarios are missing. Industry research shows that testers spend 30% to 50% of their time just preparing test data. That isn’t quality assurance. That’s operational overhead. As long as your testing department relies on 'DBA goodwill' for a refresh, true Continuous Delivery is an illusion. Real agility only happens when QA engineers get self-service access to fresh, masked, and subsetted test data via a simple portal or CI/CD pipeline integration. How many hours is your team losing this week searching for usable data?

When a database refresh takes 4 to 6 hours, something deeply human happens: teams simply stop doing it. And right at that point, an invisible drain opens in your IT budget. In our latest research, we calculated the true cost of neglected Test Data Management for a 30-person engineering team. Even with extremely conservative assumptions, we uncovered a shocking outcome: $421,800 per year in hidden waste. Where is that money going? • 45 minutes per test case lost searching for usable data in a polluted database. • 40 minutes per test case wasted manually creating test scenarios. • Rework and recovery effort caused by incomplete testing leading to bugs that leak into production. When refreshes take hours, teams improvise. They test on outdated garbage or skip chain testing entirely. The solution isn't more manual labor; it's reducing the data load. By switching from massive production copies to lean, representative subsets, you bring refresh times down from hours to 20 minutes. That single shift immediately removes $350,000+ in annual waste from your operation. I wrote this white paper about it and I'd love to hear what you think. Link in first comment.

🏛️ Et si on parlait enfin de GRU à travers des cas d'usage concrets ? Depuis plusieurs années, j'échange avec des Départements confrontés aux mêmes enjeux : centraliser les demandes, mieux orienter les usagers, simplifier le travail des agents et disposer d'une vision globale des parcours. C'est pourquoi j'ai le plaisir d'animer une série de webinars dédiés aux cas d'usage GRU dans les Départements. À chaque session, une démonstration concrète, des exemples réels et des réponses opérationnelles aux enjeux des collectivités : 🏫 Simplifier la relation avec les collèges 🤝 Améliorer l'accueil et l'orientation dans l'action sociale 👶 Fluidifier le traitement des demandes PMI 🔎 Construire une vision 360° de l'usager 🏛️ Simplifier les échanges entre le Département et les élus L'objectif n'est pas de parler théorie, mais de montrer comment ces enjeux sont traités concrètement sur le terrain. 📅 1 vendredi = 1 cas d'usage 👉 Choisissez les thématiques qui vous intéressent et rejoignez-nous : https://lnkd.in/etyQUqa9 #GRU #RelationUsager #Départements #CollectivitésTerritoriales #TransformationNumérique #ExpérienceUsager #ForumGRU

Today, we gave $1,000,000 to Open Source maintainers. Our reason is simple: support the builders of open source software we use and in general promote the community of open builders. Thank you to every maintainer. We appreciate your work Sentry. 💜 Read about where exactly the money went here: https://lnkd.in/gZNPZQk4

One thing I find fascinating about Spark optimization is how much the nuances matter. ⠀ In a recent case I worked on, we were analyzing a pipeline with low utilization, but unclear root cause. At first glance, the average task duration looked fine. But looking deeper, the average task ran for 27 seconds, while the slowest task took over 3.5 minutes. ⠀ 𝗙𝗼𝗿 𝗼𝘃𝗲𝗿 𝟮 𝗺𝗼𝗻𝘁𝗵𝘀, 𝘀𝗸𝗲𝘄 𝗮𝗰𝗰𝗼𝘂𝗻𝘁𝗲𝗱 𝗳𝗼𝗿 𝟱𝟰% 𝗼𝗳 𝘁𝗵𝗲 𝗷𝗼𝗯 𝗲𝘅𝗲𝗰𝘂𝘁𝗶𝗼𝗻 𝘁𝗶𝗺𝗲! ⠀ Detecting skew wasn't the hard part. We needed to find exactly where it occurred, why, and what key was responsible for the imbalance. ⠀ So we went down the rabbit hole, following the execution path from stages → tasks → shuffle partitions → partitioning keys – which uncovered the culprit key (𝘮𝘢𝘳𝘬𝘦𝘵_𝘪𝘥 = 2147 if you're interested). ⠀ From there, the fix was clear – salting the hot key to spread the work across partitions. ⠀ 𝗧𝗵𝗲 𝗿𝗲𝘀𝘂𝗹𝘁? 𝟮𝟰% 𝗳𝗮𝘀𝘁𝗲𝗿 𝘀𝘁𝗮𝗴𝗲 𝗿𝘂𝗻𝘁𝗶𝗺𝗲 𝗮𝗻𝗱 $𝟯𝟲𝗞 𝗮𝗻𝗻𝘂𝗮𝗹𝗶𝘇𝗲𝗱 𝘀𝗮𝘃𝗶𝗻𝗴𝘀 𝗳𝗿𝗼𝗺 𝘁𝗵𝗶𝘀 𝗷𝗼𝗯 𝗮𝗹𝗼𝗻𝗲. ⠀ This is exactly the kind of investigation we've built definity to do agentically, out of the box, so you don’t have to spend days in Spark UI, reviewing code, and re-running jobs, again and again. ⠀ In our webinar with Data Engineer Things, Roy Daniel and I will walk through a framework for optimizing Spark across the stack, using real production cases. ⠀ Stop Optimizing the Wrong Part of Your Spark Job ↓ https://lnkd.in/d5A2Pmyg

The secret is out, Web Data API is live, and it's how your agents will connect to the live web from now on 🤖 And with our latest addition to Decodo's product line, you're getting four endpoints under a single API key. Scrape pulls clean content from a single page, even behind anti-bot protection. Search returns live web results. Map discovers a site's structure, surfacing up to 147K URLs in a single call. Crawl follows links and pulls up to 10K pages in one job. Results can come back in HTML, Markdown, XHR, plus JSON, CSV, PNG where supported, so you can feed them straight into a tool call or a RAG pipeline. Under the Web Data API hood, requests run through 125M+ residential, ISP, mobile, and datacenter IPs from 195+ countries. And the best part is that we're giving up to 10K monthly requests completely free. Learn more and activate your free starter plan 👇 https://lnkd.in/dcuAzV-v

Introducing Decodo Web Data API. Depending on which source you use, AI agents are either already the biggest consumers of web data or are about to become so. Our new solution brings 4 endpoints for AI to use. Search finds live results, scrape pulls clean content from a page even behind anti-bot protection, map lays out a sites structure, and crawl collects up to 10K pages in one job. Output comes in formats agents can work with right away, including Markdown, and every request runs through our network of 125M+ IPs across 195+ countries. Let your AI agents access the web with up to 10K free monthly requests: https://lnkd.in/dRG5WEdb

Invigorated for the finish of 2026! Q4 and 2027 are packed with expanding value for GRAX's customers - and I can't wait to deliver it. This week, we gathered for QBR to celebrate the goals we met, and collaborate for next year. GRAX becoming part of Ascendx has infused so much added talent and resources - to continue innovating on the most flexible, agile, secure data & AI platform on the market. Thank you Jason Wilburn Joe Gaska Clinton Ford Stuart Newton Diego Bellalta Blaise Roulet Ufuk Civilo Bertan Ilbak Yehonatan Hahn for investing so strategically in our platform & GTM team...so I have the privilege of saying to the market... ...your full enterprise data-history is unified & ready already - just ask it questions! #enterprisedata

Anthropic just made it easier to create a dashboard, and I think that’s good news for business intelligence. With Claude Dashboards, people can ask questions about their business data and get visualizations back without writing SQL or building charts by hand. Yes, I’ve said before that dashboards are dead. Guilty as charged. The dashboard was never the end goal. People need answers they can trust and act on, and making those answers faster to produce doesn’t make them more reliable. Say revenue is down 8% in EMEA. Is it one product or all of them? Lost customers or smaller orders? And when sales and finance ask about “revenue,” are they even using the same definition? That first chart is a starting point. The real work is figuring out why. If sales and finance start from different versions of revenue, AI won’t sort that out for them. It just gets them to different answers faster. This is exactly where Qlik is focused as BI evolves. We're investing in both sides of that equation: the trusted data, quality, and governance behind the answers, and the associative analytics that help people investigate what those answers mean. Our new native connector in Claude is another step in bringing Qlik's capabilities into the AI experiences customers choose. The future of BI isn't about choosing between AI assistants and enterprise analytics. It's about bringing them together, so customers can get the speed of AI with the depth and trust of Qlik, building on the investments they've already made. I’ve written more about why the next question matters more than the first answer. Link in the comments.

How much time does your sales team spend compiling account research and prepping for sales calls? Join me to see what we've built at ThoughtSpot that saves our sellers 8 hours per week - and learn how you can build it too. Join us Oct 22nd virtually to learn more about ThoughtSpot's AgentSpot. https://lnkd.in/gykieQ4Z

I had a really great chat with Gergely Orosz on The Pragmatic Engineer about GIMP, Gmail, Colossus, erasure coding, B-trees, Swiss Tables, consensus replication, data storage vs. databases, and, of course, AI. We got into why automatic sharding matters, how strong consistency can simplify application development, and how coding agents have changed the pace and shape of my own work. As usual, the interesting parts are in the tradeoffs. We put together a recap on the blog: https://lnkd.in/gEj5JEiY Or tune into the full episode: https://lnkd.in/gCjKwQss

Great conversation with Klarna about how they're using Fullstory behavioral data to build a Buy Now, Pay Later (BNPL) checkout experience people trust. Fullstory's new survey shows why that matters: → 48% of consumers use BNPL at least once a week → 79% say they abandon a BNPL checkout at least sometimes → The #1 reason is hidden fees and unclear payment terms 81% of consumers call BNPL convenient. Most of them also abandon it. That gap isn't a payments problem; it's a UX problem, and you only close it if you can see how people actually experience your checkout, not by guessing. 🎥 https://lnkd.in/ejrWrMyp