<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Sergi Gomez on Medium]]></title>
        <description><![CDATA[Stories by Sergi Gomez on Medium]]></description>
        <link>https://medium.com/@sergi.saivo?source=rss-dc2850bf457c------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*vYtXPD6vd5kP6wQwbldsdw.jpeg</url>
            <title>Stories by Sergi Gomez on Medium</title>
            <link>https://medium.com/@sergi.saivo?source=rss-dc2850bf457c------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 20:39:51 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@sergi.saivo/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[7 highlights from 2023: Data in startups]]></title>
            <link>https://medium.com/@sergi.saivo/7-highlights-from-2023-data-in-startups-dc618a6a20b4?source=rss-dc2850bf457c------2</link>
            <guid isPermaLink="false">https://medium.com/p/dc618a6a20b4</guid>
            <dc:creator><![CDATA[Sergi Gomez]]></dc:creator>
            <pubDate>Thu, 04 Jan 2024 10:02:46 GMT</pubDate>
            <atom:updated>2024-01-04T10:02:46.649Z</atom:updated>
            <content:encoded><![CDATA[<p>I was preparing a year-end review session with my team at <a href="https://www.saivo.com/">Saivo</a> and went through all my notes from my conversations with startup founders and investors throughout 2023. Here are the trends that I distilled from all those interactions.</p><h3><strong>1. Primary needs remain unmet</strong></h3><p>Despite the excitement around Generative AI, many startups and growing companies are still grappling with the same fundamental needs in analytics:</p><ul><li><strong>(A) Full business visibility</strong>: Even the most fundamental business metrics that give us clarity on customer engagement, sales conversation rates, unit economics,… can be very challenging to obtain as we have to merge multiple data sources. I’ve often heard startup CEOs lament, “I simply don’t know how many new active users we got this month”.</li><li><strong>(B) Trust in the data</strong>: Metrics that we get off the shelf from apps often lack credibility. Many have expressed concerns like, “I don’t trust the MRR shown in my payments app”. Since they are black boxes, we don’t know what assumptions and business rules they applied to get to that number. Instead, companies should get the raw data and create their own data models that accurately reflect their business logic and semantics.</li><li><strong>(C) Self-service in data consumption</strong>: “We empower your business teams to autonomously answer questions and get insights from data”. Despite these BI platforms’ promises, I still see business teams depending heavily on tech and data teams for answering ad-hoc questions, exploring the data, creating reports, etc. This creates lots of inefficiencies that prevent organizations from unleashing the full potential in their data — More on that in point 2 below.</li></ul><h3>2. Self-service analytics: We are not there yet</h3><p>LLMs are pushing self-service to new heights. We can now ask a question in natural language, and an AI-powered apps create the query for us, run the query against our database, return the resulting values of that query, and can even create nice visualizations. I am seeing many players in the BI space building a conversational UI/UX, from new and modern ones such as <a href="https://www.zenlytic.com/">Zenlytic</a> to bigger ones such as <a href="https://www.thoughtspot.com/">Thoughspot</a>.</p><p>If we only use LLMs to write queries against a badly designed data schema, or if we don’t provide the right context, these models will generate inaccurate and inconsistent results. AI needs to be paired with the following pieces:</p><ul><li>(A)<strong> “Semantic LLMs”</strong>: This year has confirmed the necessity of the Semantic Layer as an intermediate layer between the database and the LLMs. Put simply, the Semantic Layer gives the AI model the context and the guardrails required to ensure accurate and consistent results — More on the Semantic Layer in point 3 below.</li><li>(B) <strong>The right data modeling: </strong>The hard problem in data is still to create foundation datasets of your business that integrate various data sources and are flexible enough so data consumers (or LLMs) can calculate any metric and run any type of analysis on top. You can have a very powerful AI model wrapped in a user-friendly UI to query the data, but if your data schema is wrong, you will not get the numbers accurate (“Garbage In, Garbage Out”).</li></ul><h3>3. Semantic Layer: Slower adoption than expected</h3><p>If you want an intro to the Semantic Layer, I strongly recommend <a href="https://davidsj.substack.com/p/semantic-superiority-part-1">this series</a> of articles by David Jayatillake.</p><p>In 2023, at Saivo we extensively tested Semantic Layer implementations like <a href="https://cube.dev/">Cube</a>’s and <a href="https://www.getdbt.com/">dbt</a>’s. Our preference is increasingly towards dbt-MetricFlow.</p><p>As we implemented the dbt Semantic Layer in many projects, we saw the huge potential (in autonomy, flexibility, and scalability in analytics), and we were very hopeful about a quick adoption in the market.</p><p>Yet, when I talked to data leaders, even if they agreed that the centralized source of truth for metrics is very powerful, the concept of the Semantic Layer was unknown to most of them. We are at the Early Adopters phase in the S-curve. However, for the Semantic Layer to gain broader adoption, two things need to happen:</p><ul><li><strong>(A) Compatibility with first-class BIs: </strong>The most common feedback is, “I like the idea of the semantic layer, but I will not invest in it until it works seamlessly with our existing BI” — More on BIs in point 4 below.</li><li><strong>(B) Make the Semantic Layer easier to work with</strong>. There is too much heavy lifting (YML files, branching, CI/CD…). We need to have greater agility when working on the semantic layer (editing metrics, adding dimensions, etc.) without compromising on control.</li></ul><h3>4. BI platforms: Established players remain dominant</h3><p>In my discussions, about 6 out of 10 companies using a BI tool are on Power BI. About 3 out of 10 use other major BIs (e.g., Tableau or Looker), and the remaining 10% use lower-end or modern platforms such as Data Studio or Metabase.</p><p>Even if new entrants in the market offer flashy features, customers have invested so much in creating the assets on the BI (charts, dashboards,…) and training their colleagues to adopt the tool, this makes the switching cost incredibly high. The only driver to switch is cutting costs — more on that in point 5 below.</p><h3>5. Focus on profitability</h3><p>2023 has been a difficult year for fundraising, especially in the growth stage. This has led to tighter budgets, especially in data initiatives which are a second-order need for the business.</p><p>When I speak with decision-makers, conversations often echo sentiments like, “This is not a priority for us” or “We need to bring the costs of our data infrastructure down”. They don’t need vitamins, they need painkillers, understandably so.</p><p>The macroeconomic indicators projected for 2024, such as interest rates staying almost flat, don’t suggest significant optimism for a substantial upturn in the venture market. Consequently, I expect continued focus on unit economics and profitability, which means a strong focus on ROI for everything related to data. Will this scenario lead to consolidation in the market, as many people say? Maybe, but I hold a nuanced view on this — More on that in point 6. In either case, vendors and service providers need to be ready to demonstrate tangible and high value to businesses.</p><p>Interestingly though, the current climate may have a counter-effect. Startups that want to raise have to show strong metrics, maintain meticulous reporting, and illustrate their unit economics with great clarity and accuracy. When I chat with partners at VC firms, a recurring theme starts to emerge: “We need our portfolio companies to be super diligent and effective in their reporting”. In this situation, data becomes a painkiller…</p><h3>6. Consolidation of the Data Stack</h3><p>As Jim Barksdale famously <a href="https://hbr.org/2014/06/how-to-succeed-in-business-by-bundling-and-unbundling">said</a>, “There are only two ways to make money in business: One is to bundle; the other is unbundle.”</p><p>The question in the data industry is: where do greater opportunities lie? Is it in building “all in one” data platforms (as Microsoft is trying to do with <a href="https://azure.microsoft.com/en-us/blog/introducing-microsoft-fabric-data-analytics-for-the-era-of-ai/">Fabric</a>) or in fostering “best of breed” yet narrowly focused SaaS solutions? Many data leaders have offered insightful perspectives on this debate. I recommend <a href="https://davidsj.substack.com/p/consolidation-by-other-means?sd=pf">this</a> great article by David Jayatillake.</p><p>From my conversations throughout 2023 with startups and small organizations, I see a preference for more integrated, bundled solutions. This trend resonates with the economic realities discussed in point 5 — bear markets and budget constraints.</p><p>While I am certainly not qualified to make predictions on the market dynamics, my opinion is that the modern data stack needs to be simplified, especially for small organizations. If you ask me, I would advocate for a leaner data stack like the following:</p><ol><li>Airbyte or Fivetran for data integration.</li><li>Snowflake or BigQuery as the Data Warehouse.</li><li>dbt Cloud for development, pipeline orchestration, CI/CD, documentation, and observability.</li><li>BI tool of your choice.</li></ol><p>You don’t need more than that.</p><h3>7. Increased focus on data governance</h3><p>Earlier in 2023, GitHub CEO <a href="https://techcrunch.com/2023/06/27/github-ceo-ai-and-software-development-are-now-inextricably-linked/">said</a> that “AI is lowering the cost of building great software”. I foresee this impact in the data space as well. The cost of creating data assets (metrics, charts, dashboards,…) will be decreasing rapidly with Gen AI and there will be an increasing number of data artifacts within organizations. This is undoubtedly beneficial, but it’s not without its challenges.</p><p>Imagine the following scenario: More people creating new metrics (some of them already existed), creating dashboards (that are rarely consumed),… This situation may lead to a certain degree of chaos in the organization. So, what’s the solution? Like it or not, it seems inevitable that governance will become necessary. This idea might seem contrarian with the ethos of agile, ‘move fast and break things’ startups. However, in my conversations in 2023, there is a growing call for this need, “We need to bring more control and better collaboration in our data analytics”.</p><p>Yet, we need to approach data governance more efficiently, without burdening startups with costly subscriptions for specialized tools. The solution, I believe, lies in harnessing the very technology that’s driving this problem — AI itself.</p><p>Looking ahead to 2024, it’s clear we’re all in for a challenging year in the data industry, especially for startups. We’ll need to be flexible, think creatively, and keep pushing forward. I’m really interested to see how we all, from small startups to established companies, tackle these challenges and keep growing in the data world.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=dc618a6a20b4" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[10 insights from Coalesce, the Analytics Engineering conference by dbt]]></title>
            <link>https://medium.com/@sergi.saivo/10-insights-from-coalesce-the-analytics-engineering-conference-by-dbt-8a8267edb40b?source=rss-dc2850bf457c------2</link>
            <guid isPermaLink="false">https://medium.com/p/8a8267edb40b</guid>
            <category><![CDATA[modern-data-stack]]></category>
            <category><![CDATA[dbt-labs]]></category>
            <category><![CDATA[data-analytics-tools]]></category>
            <category><![CDATA[analytics-engineering]]></category>
            <category><![CDATA[analytics]]></category>
            <dc:creator><![CDATA[Sergi Gomez]]></dc:creator>
            <pubDate>Wed, 26 Oct 2022 08:55:15 GMT</pubDate>
            <atom:updated>2022-11-06T20:27:10.648Z</atom:updated>
            <content:encoded><![CDATA[<p>Last week, <a href="https://www.getdbt.com/"><strong>dbt Labs</strong></a><strong> </strong>hosted Coalesce, arguably the most relevant event for data people these days. I attended the New Orleans edition on behalf of <a href="https://www.saivo.com/"><strong>Saivo</strong></a>, where I could meet our partners and talk to lots of people who are at the forefront of what I believe is the new wave of data analytics.</p><p>All in all, we all left Coalesce super energized. I have summarized below what I think are the 10 most relevant insights, to business and data leaders alike. I am sharing them with the aim to help those who couldn’t attend the event in person extract most of its value too.</p><p>Please note that, although these insights aim to reflect what I saw, heard, and felt, they are my own and I don’t speak on behalf of any institution or any person.</p><h4><strong>1. Most tools are empowering and amplifying the role of the analytics engineer</strong></h4><p>dbt has driven the inception of a new role in the data world, the <a href="https://www.getdbt.com/blog/we-the-purple-people/"><strong>Analytics Engineer</strong></a><strong>. </strong>In a nutshell, this role brings the best of software engineering to the data analyst. The most notable change is that data modeling is now centrally governed and version-controlled, and the same shift will happen with the <a href="https://www.getdbt.com/blog/dbt-semantic-layer/"><strong>definition of metrics </strong></a>too.</p><p>Once this model has been validated, dbt Labs, together with many other players in the ecosystem, are focused on providing tools to empower this role. The value comes in the form of higher efficiency and productivity (see <a href="https://www.selectstar.com/">Selectstar</a>, for example, where you can automate data discovery) and with higher ownership downstream in the value chain (with the modeling of the semantic layer).</p><p>However, this comes with the risk of taking freedom and flexibility from domain experts, those who don’t know (or don’t have time to use) the analytics engineering workflow and tools. Some responsibility (eg how a given metric is calculated) will be handed over to the analytics engineer, or whoever from the data team can work with these tools. I think <strong>it’s crucial to ensure the right governance is in place and to make sure the domain experts retain the flexibility they need </strong>while they leverage the advantages provided by these tools.</p><h4>2. Collaboration is the core feature of many new products</h4><p>Partly to prevent the risk presented above, new players do also acknowledge the importance of collaboration between the data team and subject matter experts and are building products to keep the latter in the loop.</p><p>These tools use different formats to enable collaboration between those who transform the data and those who are not SQL-conversant. For instance, <a href="https://count.co/"><strong>Count</strong></a> brings data teams and business together in a beautiful canvas-like format that, as the CEO told me at Coalesce, “blends the power of whiteboards with data visualization”.</p><p>Some of these companies will undoubtedly try to leverage the trend that we are seeing with the <a href="https://docs.getdbt.com/blog/understanding-the-components-of-the-dbt-semantic-layer"><strong>Semantic Layer</strong></a>.</p><h4>3. The rise of data exploration tools</h4><p>As a consequence of the above, there is a clearer distinction between traditional BI tools, such as Tableau or Superset, and others more oriented to data exploration. We have mentioned <a href="https://count.co/"><strong>Count</strong></a>, and other players are enhancing data exploration through a notebook-like format, such as <a href="https://hex.tech/"><strong>Hex</strong></a> or <a href="https://www.husprey.com/"><strong>Husprey</strong></a>.</p><p>Some of these platforms also allow users to create nice charts and dashboards, but their greatest added value lies in helping analysts dive into the data and presenting the results of the exploration. Therefore, they are complementary to traditional BI platforms. As a matter of fact, I see many data teams with a stack that includes both a data exploration tool, for first analyses and draft results, along with a reporting-focused BI.</p><h4>4. Some BIs are embracing the semantic layer</h4><p>On the one hand, <a href="https://www.looker.com/"><strong>Looker</strong></a>, with LookML, already provides the semantic layer. It’s used by many data teams and they seem to be quite happy with it. It will be interesting to see how the LookML traction evolves as the dbt Semantic Layer gains adoption, but I could feel the Looker users don&#39;t plan to switch any time soon, mainly because of the high switching cost to dbt (maybe we should build something to migrate LookML to dbt code…?).</p><p>On the other hand, we were very excited to see so much interest in open-source alternatives, namely <a href="https://www.lightdash.com/"><strong>Lightdash</strong></a>. <a href="https://techcrunch.com/2022/10/06/lightdash-takes-on-looker-with-an-open-source-bi-platform-built-for-dbt/"><strong>Lightdash is set to become the BI platform for the dbt community</strong></a>, and Coalesce definitely confirmed this great momentum to me.</p><p>BI solutions that integrate the semantic layer allow data analysts to define the business logic that is later reflected in the metrics and dimensions. Interestingly, when modeling the business logic, there is some overlap, in terms of workflow and functionality, with tools focused on data exploration that we talked about in our previous insight.</p><p>Other tools are more focused on the definition and management of the semantic layer itself, such as <a href="https://transform.co/"><strong>Transform</strong></a>, a metric platform that helps business and technical teams speak the same language and collaborate.</p><h4>5. Complexity and saturation in the tooling landscape</h4><p>During a Q&amp;A session with the dbt Labs leadership team, Tristan Handy, dbt Lab’s CEO and founder, said:</p><blockquote>“We are moving into the deployment phase, where we have to figure out what to do with all the toys we have available”.</blockquote><p>I could confirm this when chatting with many data teams throughout the event. It was common to hear something like “we are overwhelmed with all the new tools that we find every week”.</p><p>Along these lines, they also said that it’s difficult to switch from one tool to another, especially when it requires adoption by non-technical stakeholders. Plus, the procurement process of those tools sometimes leads to internal friction within the organization (“are we buying another SaaS tool?”).</p><h4>6. Either consolidation or better integration</h4><p>Building upon the claim raised by Tristan, Nicholas Erdenberge, dbt Lab’s VP of sales, added:</p><blockquote>“We will see consolidation in the stack in the next years”.</blockquote><p>And, once again, I could confirm this in my discussions with analytics engineers, who told me that “we are looking forward to the day that the data stack consolidates into a smaller set of standard tools”. Unfortunately, this scenario would result in many of the tools not making it to the “standard data stack”.</p><p>Yet, I personally bet on a more optimistic scenario. <strong>If we enhance the integration from one tool to another, we can create powerful synergies and a more seamless experience across these tools</strong>. This will make it easier for users to adopt and switch among them. We’ll end up with a less fragmented landscape where tools “communicate” better with each other.</p><h4>7. Open-core, the approach taken by many players</h4><p>One of the three core principles of dbt states:</p><blockquote>“We want to create more value than we extract”.</blockquote><p>And you can truly feel this from its community. Many people claim the analytics engineering community is the most open and friendliest of all, and I don’t think it’s an overstatement.</p><p>Beyond its philosophical meaning, and just as Max Beauchemin (<a href="https://superset.apache.org/"><strong>Superset</strong></a>’s Founder and CEO) said, open source is a superior approach to developing software, also from a practical standpoint.</p><p>In fact, dbt and many other players in the ecosystem are taking the “open-core” approach: They make the software open-source, which is maintained and developed mainly by their team, and they make money off consulting, hosting, and offering “Enterprise” features for those technologies.</p><p>The following are some really great open source technologies across the stack:</p><ul><li><a href="https://airbyte.com/"><strong>Airbyte</strong></a> for data integration</li><li><a href="https://airflow.apache.org/"><strong>Airflow</strong></a> for data orchestration</li><li><a href="https://docs.getdbt.com/docs/get-started/getting-started-dbt-core"><strong>dbt Core</strong></a> for data transformation</li><li><a href="https://transform.co/metricflow/"><strong>Metric Flow</strong></a> for the semantic layer</li><li><a href="https://www.lightdash.com/"><strong>Ligthdash</strong></a> and <a href="https://superset.apache.org/"><strong>Superset</strong></a> for BI</li><li><a href="https://greatexpectations.io/"><strong>Great Expectations</strong></a> for quality assurance</li></ul><p>These players, among others, are democratizing the access to the modern data stack and thus the amount of value they create is immense. While they need to capture part of this value (aka monetization), this strategy allows everyone of us to use (and contribute to) some best-in-class tools for free.</p><h4>8. The adoption of dbt is in its early days</h4><p>The momentum that we are seeing with the dbt adoption is impressive. Everyone I speak with, I keep hearing that “we may have doubts about some pieces of our stack, but there is one thing we are 100% certain about: we’ll use dbt for data transformation”.</p><p>Yet, dbt is still new to many data folks. At Saivo, for instance, our backend generates the dbt models automatically based on the metrics created on a no-code interface (btw, we’re super excited to start supporting dbt metrics!). Then, most of our customers hire their first data analyst at some point. More often than not, this person hasn’t heard of dbt and might be a bit reluctant about it “can’t I just write the queries on the data warehouse?” Well, <em>every single person </em>(data analyst, data engineer, whoever) that we have onboarded to dbt (so she can extend or edit the dbt models generated by Saivo), has expressed a significant gain in efficiency and confidence when preparing the data using dbt.</p><blockquote>“In the world, there are 50k dbt users, and 12.5 million people who know SQL”.</blockquote><p>This was shared by dbt at Coalesce. And, by the way, out of these 50k users, there are 16k companies already using dbt, and 3k of them are on dbt Cloud. That is, almost 20% of the companies are already on dbt Cloud. This tells us about the fast growth of dbt Cloud, and it will continue this way in the next few months, as dbt Labs is giving it a strong commercial push.</p><p>At Coalesce, I could confirm, in every interaction that I had, that dbt’s user base is incredibly engaged and passionate about it. And, while there is a long way to go for dbt, we all know that “it’s better to have 100 users that love your product, than 10000 that somewhat like it”.</p><h4>9. Service providers will also leverage dbt’s momentum</h4><p>This rapid and strong adoption of dbt is not only leveraged by software companies, but also by service providers, such as analytics consulting firms.</p><p>In fact, dbt already has ~2000 partners around the world. These partners deploy, maintain, or simply use dbt in the services they provide, and they also offer other services such as training or certifications.</p><p>At Coalesce, I could talk to quite a few analytics consulting firms. Although they typically offer end-to-end services, they see dbt as the cornerstone of their service offering.</p><p>In North America, most of these firms’ clients already knew about dbt and wanted to accelerate the transition into the modern data stack. While it’s definitely easier to sell these types of services in a more mature market, I can see a great opportunity for the European counterparts in the next few years.</p><h4>10. Lack of trust is still an important challenge</h4><p>A common theme that I could see and hear in many talks is that there is still one important challenge for data teams, which is the lack of trust from business stakeholders in the data.</p><p>I was struck by the following quote presented on a talk I attended:</p><blockquote>“73% of Product, Go-to-Market, &amp; Finance teams don’t feel they can trust data generated by their data teams”</blockquote><p>I consider this one of the most important challenges to be tackled, especially for those who are building products or services to help people use data for decision-making.</p><p>Take dbt Lab for instance. Its mission is to empower data practitioners to create and disseminate organizational knowledge. And, as they put it, there are three requirements to disseminate knowledge:</p><ul><li>Make it easy to create a new piece of knowledge</li><li>Make this knowledge widely accessible</li><li>Establish trust in this knowledge</li></ul><p>So, if we can’t get the trust of the people who look at the data, we’ll not be able to disseminate knowledge and our data initiatives won’t succeed.</p><p>True, there are many players working very hard on this problem, mainly through products that offer enhanced data observability and data discovery. However, I feel that there’s still some progress to be made and that more efforts should be allocated to this.</p><p>That’s it! I hope these insights are helpful. Don’t hesitate to get in touch on <a href="https://www.linkedin.com/in/sergigomezpalleja/">LinkedIn</a> if you have any questions, want me to elaborate further on any of the points above, or if you simply want to have a chat.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3OFFrYqYXwuTYaWlot7IKg.jpeg" /><figcaption>Tristan Handy, dbt Lab’s founder and CEO, at the main Coalesce’s keynote on October 2022.</figcaption></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8a8267edb40b" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>