October 2012 Summaries
5 posts from Zapier
Filter
Month:
Year:
Post Summaries
Back to Blog
Zapier now supports over 100 public applications on its platform, automating millions of tasks every month for thousands of people worldwide. The company was founded in 2011 by Bryan and Plushcap (me), with the original idea born out of a conversation between us. We bootstrapped our way to sales, initially selling beta invites for $100 each, and later gained traction enough to go full-time in January 2012. After experimenting with product-market fit, we launched a new interface that tripled engagement with our customers, and were accepted into Y Combinator in April 2012. We've since launched the developer platform, supported over 100 apps, and secured funding from Bessemer Venture Partners and other investors. With this support, Zapier will continue to grow, adding more services, features, and a team that cares about helping businesses run more efficiently.
Oct 31, 2012
1,281 words in the original blog post.
We use SQLAlchemy at Zapier to handle external database connections to various databases such as MySQL, MSSQL, PostgreSQL. Recently, we encountered a problem with our MSSQL connector where using a username or password with a plus sign (+) caused connection issues due to double URL encoding of the odbc_connect parameter in the SQLAlchemy connection string. The issue was resolved by properly URL encoding the odbc_connect parameter twice to avoid decoding errors and ensure correct password transmission. To connect to MSSQL with SQLAlchemy, we use a special connection string that includes extra variables encoded into the odbc_connect parameter. For example, our working code generates a mssql+pyodbc SQLAlchemy connection using the urllib.quote_plus function to URL encode the odbc_connect parameter twice.
Oct 31, 2012
404 words in the original blog post.
We started by combining our own needs for integrations with recent web API explosions at Startup Weekend, where we won a demo but the real work was just beginning. We bootstrapped our product by finding paying customers through support forums and building features around their requests. After a second submission to YC, we prepared extensively for an interview, pestering alumni for advice and conducting mock interviews to feel more comfortable. On interview day, we stood out by wearing matching t-shirts, but the conversation was surprisingly casual and brainstorm-like, with the partners discussing our product platform. After a short interview, we waited anxiously for the call of acceptance or rejection, which ultimately came in our favor when Paul from YC called to inform us that they would fund us.
Oct 30, 2012
645 words in the original blog post.
When should a startup consider buying a commercial solution instead of building one? It makes sense to buy when the problem being faced doesn't directly tie closely to the core value proposition, or when existing solutions aren't specific enough to meet the company's needs. On the other hand, startups should build their own solutions when the problem is closely tied to their core value proposition and there isn't a suitable commercial solution available that meets their unique requirements. In Zapier's case, they built their live chat and drip campaign tools because these were complex problems that required custom solutions to meet their specific needs, such as supporting diverse user bases and providing control over email campaigns. Ultimately, the decision to build or buy depends on the individual company's circumstances and should be made after careful consideration of both options.
Oct 14, 2012
523 words in the original blog post.
This past weekend, I had the opportunity to attend Columbia MO's second annual Startup Weekend where I teamed up with Wade to build a gag startup called PRLibs, which uses madlibs to auto-generate press releases for startups. We realized that many startup press releases sound similar and could be improved by creating an automated system. To achieve this, we used Posterous and Zapier to create a simple HTML form, send emails, and generate paragraphs from user input. Our project was successful, despite being referred to as "worst startup of all time", and demonstrated the potential for non-developers to build working demos using various tools such as Webhooks app, Kickoff Labs Landing pages, Balsamiq, WooThemes, Posterous, and Tumblr. The experience showcased that with the right combination of tools and a bit of creativity, anyone can turn their idea into a working prototype.
Oct 10, 2012
422 words in the original blog post.