Merge's Unified API Bet in the AI Era | E2033

26 Oct 2024 (27 days ago)
Merge's Unified API Bet in the AI Era | E2033

Alex kicks off the show with guest Shensi Ding of Merge. (0s)

  • Startups often approach Merge to integrate their services, as it provides free distribution and helps create an ecosystem around their products, making it easier for them to compete with larger companies like HubSpot or Salesforce (16s).
  • Merge offers a unified API that provides access to multiple other APIs, simplifying the process of working with various software products, data types, and points of integration (2m23s).
  • APIs, or application programming interfaces, have become a crucial part of the software industry, allowing different pieces of code to interact with each other, and companies like Twilio have successfully leveraged APIs to offer their products with usage-based pricing (1m44s).
  • The number of APIs has increased significantly, leading to complexity in working with multiple software products, data types, and points of integration, which Merge aims to address with its unified API (2m15s).
  • Merge's co-founder, Shiny Ding, believes that offering an API is now a standard requirement for companies, and Merge's API provides a solution for startups and companies looking to integrate their services with others (2m27s).
  • The conversation with Shiny Ding aims to discuss the current state of APIs, as the topic was widely discussed in 2021 and 2022 but has since faded into the background, despite their continued importance (3m7s).
  • The founding story of Merge highlights the problem of API complexity and the need for a unified solution, which is what Merge aims to provide (3m28s).
  • The company was founded by two individuals who met while studying computer science in New York City, where they were in the same classes and worked on group projects together. (3m46s)
  • The co-founders, who were also part of the engineering student council, developed a strong friendship throughout college, with one of the founders noting that his counterpart would always follow through on commitments. (4m11s)
  • After college, the co-founders went down different paths, with one becoming a skilled software engineer and the other pursuing a career in investment banking in New York City to learn about the business side of companies. (4m46s)
  • The individual who pursued investment banking spent two years at Credit Suisse, where they learned to read financial statements and understand businesses, but eventually moved to San Francisco to rejoin the tech industry. (4m51s)
  • In San Francisco, the individual joined Silver Lake's growth equity fund and reconnected with their co-founder, who helped them settle into the city, but they soon realized that investing was not close enough to operating a company. (5m15s)

Shensi's journey from finance to tech and the importance of integrations (5m18s)

  • The role of Chief of Staff is described as being the CEO's right-hand person, responsible for ensuring that tasks are completed and the company operates smoothly. It is often seen in companies with operational challenges, as it helps guarantee execution. (5m51s)
  • The Chief of Staff role involves a variety of tasks, including documentation, recruiting, and ensuring that the CEO's directives are carried out. It is likened to being a COO for the CEO, with a focus on execution rather than strategy. (6m22s)
  • The speaker gained significant exposure to the CEO, Tim, who was knowledgeable in politics and cyber warfare, providing a valuable learning experience about the intersection of technology and global dynamics. (7m2s)
  • During the speaker's time at Expanse, it was observed that integrations were a significant challenge for the business, often causing the company to lose deals despite having a superior product. Competitors with less high-quality products were winning based on features related to integrations. (7m41s)
  • Integrations became a major concern when the sales team started demanding more of them, as the lack of integrations was hindering the company's ability to close deals and ultimately affecting revenue (8m21s).
  • Building integrations was a costly endeavor, requiring the hiring of expensive engineers, which would significantly impact operating expenses, and even moving to lower-cost areas would not alleviate the issue (8m41s).
  • Maintaining integrations was also a challenge, as it required a permanent team of engineers and became a permanent line item in the company's expenses (8m53s).
  • The co-founder, Gil, had previously experienced the difficulties of building integrations firsthand while working at a recruiting tech company, where he had to build integrations by himself due to the company's limited resources (9m30s).
  • Gil's experience revealed that building integrations involved several challenges, including obtaining access to APIs, getting partnerships, and navigating complex workflows, which could lead to costly mistakes and delays (9m47s).
  • In one instance, the product manager and design team spent three weeks building an integration into the wrong API, resulting in wasted engineering time and significant financial losses (10m7s).

The challenges of building and maintaining integrations (10m20s)

  • A lean Series A team spent three weeks building an integration, which resulted in a significant amount of money and opportunity cost, and it took six weeks to complete the full end-to-end correct integration, despite the team's hard work and efficiency (10m20s).
  • Even after the integration went live, the engineers had to dedicate time to helping answer customer success questions, as customer success team members were unable to resolve issues on their own (10m38s).
  • Customer success team members received questions from recruiters, such as issues with data syncing, missing candidate information, and resume uploading problems, which they would then escalate to the engineering team (10m48s).
  • The engineering team had to allocate a percentage of their time to help customer support with troubleshooting, often finding that the issues were due to end-user error, such as incorrect API keys or expired keys due to employee changes (11m2s).
  • These issues were not handled by any piece of software, resulting in a significant burden on the engineering team (11m17s).

Coda. Empower your startup with Coda’s Team plan for free—get 6 months (11m19s)

  • Founders often spend too much time switching between team chat, document editors, spreadsheets, and databases, and it's time to consolidate all that knowledge into one platform (11m20s).
  • Coda is a collaborative workspace that pulls together documents, spreadsheets, databases, and apps, making it easy to learn and incredibly powerful (11m37s).
  • Coda is used to run Founder University and was also used to create a database for the top 500 private companies, which was a no-brainer due to its capabilities (11m55s).
  • Coda empowers startups to strategize, plan, and track goals effectively, and offers a limited time offer for startups to get 6 months free of their team plan at coda.io/twist (12m23s).
  • With Coda, startups can save at least a couple of hours a week per team member and feel in control, and can get started for free and get six free months of the team plan at coda.io/twist (12m33s).
  • Many startups face issues with deals going out the window due to lack of integrations, despite having a better product, and Merge has built a solution to this problem with their unified API (12m51s).

Identifying the market gap for Merge and the benefits of standardized data models (13m17s)

  • When the company was founded, APIs weren't new, but the market hadn't closed the gap in integrations, partly because popular companies like Trade, Zapier, and Rado were prevalent, and the onus of building integrations was on the buyer, not the vendor (13m23s).
  • At that time, it was common for companies to purchase multiple software solutions and then connect them using third-party integration tools, which was a pain to maintain and expensive, with costs ranging from $3 to $100,000 (13m57s).
  • Software wasn't as fragmented, and companies often purchased a whole suite of software from a single provider, reducing the need for integrations (14m27s).
  • The idea behind Merge is to have one API that brings together multiple APIs, providing a single conduit for all other APIs, with the same credentials, pagination, and rate limiting, but with different objects or data models per category (14m57s).
  • Having normalized data models allows for more total partners, as it's easier to link in new sources of information, and fundamentally, every single source of truth has a very similar data model (15m29s).
  • People need integrations to land sales, and software fragmentation means that companies have a different mix of software solutions, making it impossible to sell them something new without having every single incoming port (16m17s).
  • The unified API approach is especially relevant in the AI era, where integrations are crucial for companies to adopt new technologies and solutions [no timestamp mentioned in the text, but it's the title of the video].
  • Companies today have around 200 pieces of software that they are trying to tie together, which can be a day-to-day issue, as seen in a recent report by Vendor Curement. (16m32s)
  • Merge, with 110 people, has a ton of software providers and expects each of them to sync data with each other, which can be a huge pain, especially with millions of rows of data getting updated daily. (16m41s)
  • The expectation of buyers is high, with many vendors available, and they prefer to purchase something that seamlessly syncs with all their other vendors, eliminating data silos and ensuring consistent data across all vendors. (17m8s)
  • Merge has partnered with companies like Paylocity, and there may be tension between providing a unified API and individual companies offering their own API, potentially not wanting to be aggregated into someone else's service. (17m21s)

Partnership building, Salesforce integration, and overcoming resistance (17m39s)

  • Building partnerships requires a lot of evangelism and networking, and sometimes it's necessary to reach out to old contacts to get in touch with key people at other companies (17m39s).
  • In the early days, the company had to do whatever it took to get in front of the C-suite at a large HR platform company, which eventually became one of their best partnerships (18m7s).
  • The goal for every source of truth is to create an ecosystem around their product, similar to what Salesforce has done, making it impossible to not use their platform (18m32s).
  • Salesforce has become a critical component of many sales ops tools, and not building on top of it can make it harder for companies to succeed (18m37s).
  • Companies want to get built around them, becoming a source of truth and critical to their customers' operations, making it hard to turn them off (19m0s).
  • An example of a company that built its entire business on top of Salesforce's platform is Ino, which went public and added value to Salesforce's operations (19m6s).
  • The success of companies like Salesforce can lead to growth and expansion, such as adding more people to their team, as seen with the addition of 5,000 people to Dreamforce (19m18s).
  • Dreamforce, Salesforce's annual conference, can be a significant event, taking over venues and causing some disruption to the local community, but also providing a fun experience for attendees (19m25s).
  • The speaker has attended Dreamforce and enjoyed the experience, including a concert by Imagine Dragons, which they acknowledged was not their usual taste in music but still enjoyable (19m45s).
  • The speaker is familiar with the soundtrack from League of Legends and acknowledges that it has some great songs, including those by Imagine Dragons (20m0s).

OpenPhone - Get 20% off your first six months (20m7s)

  • OpenPhone is a modern business phone solution that works through a single, elegant app on existing phones and desktops, eliminating the need for multiple devices and apps to run a business (20m9s).
  • The app allows users to separate business and personal communications, enabling them to text and call customers and clients without using their personal phones (20m36s).
  • Over 50,000 businesses use OpenPhone, which offers a more affordable alternative to traditional phone systems that can cost $500 to $1,000 per person per year (20m43s).
  • OpenPhone costs $13 a month, and users can get an extra 20% off any plan for the first six months by visiting OpenPhone.com/Twist (20m57s).
  • Existing phone numbers can be easily ported over to OpenPhone at no extra cost, allowing for a seamless transition (21m4s).
  • Users can start their free trial and get 20% off by visiting OpenPhone.com/Twist (21m11s).

Further discussion on trust-building and integration challenges (21m16s)

  • Merge, a company, faced skepticism from other companies in 2020 due to its small size, making it difficult to establish partnerships and integrations, but this changed over time as the company grew and gained more integrations (21m17s).
  • The process of establishing partnerships was challenging, with many people doubting Merge's ability to solve their use cases and provide deep enough solutions, similar to convincing people to move from on-prem to the cloud (21m47s).
  • Merge's business model was also difficult to understand, as it wasn't transactional like other API companies, and pricing and customer support were harder to figure out (22m22s).
  • The company had to navigate a complex customer buying process, but this is also where Merge provides value by abstracting away complexity through its APIs (22m56s).
  • Merge's API abstracts away complexity by handling integrations with various providers, such as accounting and HR providers, which forms a significant part of its competitive advantage (23m0s).
  • However, the risk of copycats following up on Merge's work and replicating its integrations could potentially drain its competitive advantage, but Merge's first-mover advantage and large customer base provide a maturity curve for its integrations (23m22s).
  • Merge's extensive customer base and continuous onboarding process help the company discover more edge cases and improve its integrations, as API documentation is often outdated and response bodies can vary wildly (23m40s).
  • Merge has an advantage over in-house integrations due to its large customer base, which provides numerous data points for edge cases, making it less likely for customers to encounter bugs when onboarding to Enterprise Integration (24m5s).
  • The company's ability to see data in motion, thanks to its many pipes running between different products, raises the possibility of a data observability element (24m27s).
  • Merge is referred to as an integration observability firm, providing visibility into data flowing in and out of the platform, which is essential for B2B customers where reliability is crucial and each end-user can cost millions of dollars (24m49s).
  • The company's customer support teams primarily interact with Merge, and its support team ensures that every piece of data flowing through the platform is visible to customers (24m57s).
  • Merge's dashboard provides tooling focused on non-technical people, allowing them to view and modify API requests, scopes, and data flow, as well as search for specific information, such as user data (25m31s).
  • The platform stores information, making it easily accessible, and provides visibility into end-user errors, simulating the functionality of tools like Data Dog and Sentry for in-house integrations (25m43s).
  • Unlike in-house integrations, which require engineers to use multiple components to troubleshoot, Merge's dashboard allows customer support personnel to easily identify and resolve issues (26m0s).

Enhancing integration observability and customer support (26m4s)

  • Merge has a product called Blueprint, which allows users to crowdsource integration recommendations, utilizing AI to generate initial mappings and documentation, although it's not a full-blown product and mostly an experiment to see if customers can contribute to integrations (26m13s).
  • The main blocker for building integrations is sandbox access, not documentation, as it requires testing and modifying the UI to ensure the integration works properly (27m10s).
  • Sandbox access is hard to get due to the need for advanced partnerships, requiring the ability to issue an account to a partner for testing purposes, which can be a complex process (27m48s).
  • Some categories, such as HR Tech, are not self-served, making it difficult to obtain sandbox access, while others like ticketing and help desk are easier to access (28m6s).
  • Merge makes a big deal out of adding new integrations, which can be beneficial for startups looking to get their name and logo on Merge's site and materials (28m42s).
  • Companies, especially startups, see Merge as a distribution point, wanting to integrate with Merge to increase their visibility and reach more customers (28m56s).
  • Merge allows companies to integrate with them for free, creating an ecosystem around their product and providing free distribution, which can be especially helpful for new companies trying to compete with established vendors (29m39s).

Micro1 - Visit to get 10 free AI Interviews and 2 weeks free per hire (30m1s)

  • Micro1 is a platform that helps startups scale their product teams in days, rather than weeks or months, by handling the search process, vetting, onboarding, and paperwork (30m12s).
  • Micro1 has built an AI engine that interviews tens of thousands of engineers, designers, and AI trainers every month, allowing them to identify the top 1% of candidates (30m31s).
  • Startups can use Micro1 to accelerate their product development by finding and onboarding top talent quickly, without having to vet numerous resumes or hire recruiters (30m46s).
  • To use Micro1, startups simply need to specify their engineering needs, and Micro1 will help them find their next engineer in 48 hours or less (30m54s).
  • By visiting Micro1.ai, startups can take advantage of a special offer, which includes two weeks of free development per hire and 10 free AI interviews (31m2s).

Expanding Merge's integration list and its AI applications (31m21s)

  • Merge's primary focus is on building the most comprehensive list of integrations, with a goal of having the highest quality integrations and product, rather than charging customers for access at this point (31m23s).
  • The company has experimented with charging for access in the past, but currently prioritizes having the best product with the most integrations (31m36s).
  • Merge aims to have all the big integrations needed, but acknowledges that there's an unlimited number of integrations to build, especially with the increasing internationalization of the world (31m50s).
  • The company is getting faster and better at building integrations, focusing on creating efficiencies in testing and maintenance to move quickly and provide high-quality services to customers (32m19s).
  • Merge's operational difficulty is not just technical, but also operational, as it aims to be cross-category and international, and has built its foundation to avoid technical debt from the ground up (32m50s).
  • The company had security on its mind early on, achieving SOC 2 compliance from the start, thanks to its co-founder (33m9s).
  • Merge's AI product is designed for companies with many customers and different integrations to bring data in and make AI features for their customers, rather than for internal use cases (33m35s).
  • The demand for Merge's AI product has been prevalent, especially with the AI boom, as companies want to enhance their product insights and capabilities with customer-specific data (34m9s).
  • Many companies are being founded with a focus on AI, making it a significant trend in the industry (34m13s).
  • In the Enterprise segment, it's harder to distinguish between AI-specific companies and those that simply add AI to their products (34m19s).
  • The quality and uniqueness of data used to build AI models play a crucial role in differentiating them and making them smarter (34m34s).
  • Having more unique data inside an AI-predicated application enables it to perform better and do more (34m41s).
  • Unique data is essential for AI applications, and companies are moving towards using more AI, which may increase the total throughput of information across APIs (34m46s).
  • Merge has seen an increase in companies using multiple categories of integrations, with some using all seven categories, indicating a growing need for diverse data (35m0s).
  • The AI boom has provided a favorable business environment for Merge, as the more data available, the better the insights, and the more valuable the AI applications (35m12s).
  • The current AI trend has been beneficial for Merge's business, with the company's CRO attributing some of their success to being "lucky" (35m20s).

Discussing financial efficiency, runway management, and team dynamics (35m27s)

  • Merge has raised $74.5 million in funding, with the last round being a $55 million Series B in October 2022, and a post-money valuation of $315 million, according to Crunchbase and Pitchbook. (35m36s)
  • The company is not currently raising more money, having been efficient with their funds and not spending most of the money from their Series B, thanks in part to their Finance team, led by a CFO equivalent who is ensuring the company is being frugal, especially in the current market. (35m54s)
  • The company has a long runway with their current spending, and their Finance team is actively managing their cash and checks. (36m15s)
  • Merge has a more mature Finance operation than most companies of their size, with a CFO equivalent and a CRO, and only three rounds of capital raised. (36m26s)
  • The company's early investment in the Office of the CFO has been beneficial, with the CFO equivalent, Alexia, being hired when the company was 10 people, to handle operations and accounting tasks that the founder wasn't uniquely good at. (37m3s)
  • The CFO equivalent has been able to be revenue-generating and strategic in expanding to different products and regions, and has helped the company to be thoughtful about pricing and cost compensation. (37m25s)
  • The Finance team has expanded over time, with Alexia starting with a lean team during the Series A, and then expanding to two people after the Series A, and then to three people after the Series B, with the addition of a VP of RevOps and two RevOps team members. (37m41s)
  • The company's finance and operations team is "pretty loaded" for a Series B company, but the complexity of their operations has made it an advantage for them, with deep insights and understanding of their sales motion and pipeline. (37m59s)
  • The company's $55 million Series B round was led by Excel, and was a relatively outsized round, even for 2020. (38m21s)
  • Merge was conservative with its new investment, wanting to make strategic investments in certain areas with the $55 million raised, but not spending all of it (38m44s).
  • The company's approach to fundraising and spending was influenced by the founder's previous experience at Expanse, which was acquired for $1 billion with only 200 employees and a focus on large Enterprise customers (39m2s).
  • Merge started in June 2020, a challenging time for fundraising, and the founder knew the business would be expensive to build due to the need for technical talent in San Francisco and New York City (39m24s).
  • The founder prioritized protecting the team and avoiding layoffs by hiring slowly and ensuring everyone was highly skilled, which was made more challenging by being an in-person company (39m39s).
  • The in-person requirement limited the candidate pool and slowed down the hiring rate, but the founder respected every company's right to make their own decision on this matter (40m5s).
  • Despite the challenges, Merge managed to conserve its funds, still having a significant portion of its Series B two years after raising it, which is a rare achievement in the market (40m30s).
  • The founder attributes the company's success in managing expenses to having a close-knit team that feels comfortable calling out unnecessary investments and encouraging thoughtful decision-making (40m46s).
  • Merge has experienced significant growth, with its revenue growing by 30X in the last 12 months and now having around 16,000 self-served organizations and 400 Enterprise customers on the platform (41m22s).
  • The company has recomposed its revenue structure, shifting from primarily serving small and medium-sized businesses to now having a mix of small, medium, and large enterprises (41m37s).
  • Merge's Series B funding allowed the company to focus on investing in its future, improving product quality, and expanding its total addressable market (TAM) (42m1s).
  • The company's approach to growth has been to alternate between going deep in specific categories and then expanding broader, allowing it to balance depth and breadth without being spread too thin (42m47s).
  • Merge has benefited from being a first-mover in its space, giving it time to establish itself and grow efficiently (42m55s).
  • The company's growth strategy prioritizes efficiency and sustainability, with a focus on protecting the team and ensuring a good outcome, rather than over-fundraising and over-spending (43m7s).

Growth strategies, customer base expansion, and go-to-market experiments (43m16s)

  • The company has had a good year in 2023, doubling the number of logos and expanding product offerings, with some big logos potentially closing in Q4 that they wouldn't have had conversations with last year (43m30s).
  • If these new deals land in Q4, they will likely pull up the Average Contract Value (ACV) in the Enterprise segment, but probably not overall, due to the presence of smaller customers (43m49s).
  • The majority of Merge's revenue today comes from the Enterprise segment, which is growing a lot, and has increased its percentage of revenue significantly since the company's early days (44m15s).
  • Going upmarket has also opened up a new industry sector mix to the company, with Financial services now being a potential market, which was not possible before (44m45s).
  • The company has not considered changing its approach to cash and trying to grow faster, preferring a conservative approach to spending and ensuring a clear return on investment (45m11s).
  • Merge's go-to-market approach involves being present at industry events, meeting customers in person, and experimenting with different regions, with a focus on meeting more customers and understanding their needs (45m59s).
  • The company has found that meeting people in person helps to build understanding of their unified APIs, but there is still a lot of education required, even among people in the industry (46m27s).
  • The company's CEO often finds himself explaining what the company does, even in social situations like weddings, and values the in-person interactions as a way to build understanding and trust (46m40s).

Pricing strategies, customer value, and market trends (47m10s)

  • Merge's pricing strategy is based on the number of linked accounts, with the first tier offering three free linked accounts and subsequent tiers charging $650 for up to 10 linked accounts, with each linked account being a customer endpoint integration into the API (47m10s).
  • A linked account is a connection, such as a customer connecting their Workday instance through Merge, and the pricing is roughly 1/1,000th of the potential value for the customer (47m38s).
  • The initial pricing was based on volume, but it created complexities in the sales cycle, so Merge simplified the pricing to focus on the majority of smaller LinkedIn accounts rather than the few large instances (48m2s).
  • The current pricing is a good deal for companies selling to Enterprise customers, who can also make money from integrations, and Merge is open to adjusting the pricing in the future (48m24s).
  • The company prioritizes providing a great experience for customers and is willing to leave value on the table, but it's a challenge to decide when to adjust pricing to avoid leaving too much value on the table (48m40s).
  • Merge has a platform fee for professional and Enterprise plans, which varies based on company size, and this fee helps to offset the cost of supporting larger customers (48m51s).
  • The trend of software companies reporting poor net retention rates has not significantly impacted Merge's growth rate, as the company's pricing is based on the number of connections, and the move upmarket has helped to mitigate this issue (49m32s).
  • The net retention rate (NRR) for Merge's Enterprise segment is significantly different from its S&B segment, and the company has been fortunate in making the strategic move to focus on Enterprise customers earlier on (49m57s).

Future plans for Merge, acquisition interest, and long-term vision (50m5s)

  • Merge is expected to raise more capital and grow significantly, potentially leading to an IPO in Q2 2027, as it has good growth, product-market fit, and a solid team (50m6s).
  • The goal is for Merge to go public, as it's believed that all tech companies that go public eventually become M&A shops, strategically purchasing other companies to benefit customers (50m44s).
  • There are a few bets that Merge wants to make as a private company, which might not be well-received by the stock market, so it's essential to make these bets before going public (50m55s).
  • There is no minimum personnel requirement to go public, and companies can go public with as few as two people (51m14s).
  • Merge has not received any acquisition offers, which is surprising, especially from companies in the AI data space like Databricks (51m55s).
  • Merge is not interested in being acquired, as the opportunity for what they're building is too big, and it would be a shame to give up on their vision (52m8s).
  • Merge believes they are building the infrastructure for the next generation of the internet, emphasizing the importance of data, and doesn't think an acquisition would be worth it at this point (52m15s).

Evolution of Merge beyond API focus. (52m26s)

  • Merge aims to become the data interchange for the internet, especially on the business side, and is evolving beyond its API focus to become more agnostic about types of data in motion (52m26s).
  • The company is working on expanding its capabilities to pull and export customer data in various formats, regardless of where the data is located (52m51s).
  • Merge has already developed features such as CSV upload, on-premises integrations, and SFTP, with plans to continue expanding its offerings (52m57s).
  • The company prefers to be referred to simply as "Merge" rather than "Merge API," as it is moving beyond its API-focused roots (52m45s).
  • The goal is to provide a unified platform for data interchange, allowing customers to export their data in the format of their choice (52m54s).

Overwhelmed by Endless Content?