{"id":26004,"date":"2022-12-01T21:07:35","date_gmt":"2022-12-01T21:07:35","guid":{"rendered":"https:\/\/www.intentwise.com\/blog\/?p=26004"},"modified":"2024-02-06T05:52:15","modified_gmt":"2024-02-06T13:52:15","slug":"three-lessons-from-building-a-data-stack-on-amazons-api","status":"publish","type":"post","link":"https:\/\/www.intentwise.com\/blog\/technology\/three-lessons-from-building-a-data-stack-on-amazons-api\/","title":{"rendered":"Three Lessons From Building a Data Stack on Amazon\u2019s APIs","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"\n

As someone working with ecommerce data, you might think unlocking Amazon data is as simple as connecting to Amazon’s APIs. Unfortunately, you\u2019d be wrong. At Intentwise, we have built a complex suite of solutions\u2014Intentwise Analytics Cloud<\/a>, Intentwise Ad Optimizer<\/a>, and Intentwise DSP<\/a>\u2014to help brands, agencies, and aggregators better manage their data strategies. Through this process, and in our many, many conversations with professionals like you, we\u2019ve learned a few things. Whether you intend to build your own internal Amazon data infrastructure, or you\u2019re shopping for an off-the-shelf solution like ours, here are three important lessons we\u2019ve learned that may help guide your decision-making.<\/p>\n\n\n\n

Amazon\u2019s API connections need constant TLC.<\/h2>\n\n\n\n

You likely have a simple ask of your ecommerce data reporting system: give me accurate data at a reliable time. Integrating with Amazon’s APIs is the first, and easiest, challenge. Once you make the initial connections, you must be able to react and adapt to API changes and occasional data gaps.<\/p>\n\n\n\n

In 2022 alone, Amazon\u2019s data landscape has changed dramatically. Amazon released a whopping five versions of the Sponsored Ads API this year. Amazon Marketing Stream<\/a> and Amazon Marketing Cloud<\/a> present new challenges in integrating that data as well as an abundance of growth opportunities. Amazon released APIs for Vendor Central in March. The list goes on and on.<\/p>\n\n\n\n

While Amazon is sharing a ton of information with brands and sellers, the wealth of data requires ongoing investment to keep up with these APIs. It\u2019s crucial that your data pipelines are comprehensive. If you are building a system internally, expect constant change and new updates throughout the year\u2014it\u2019s not a simple one-and-done connection to Amazon\u2019s APIs.<\/p>\n\n\n\n

Data errors are a given. Plan for them.<\/h2>\n\n\n\n

Although a future in which data is clean, reliable, and on-time would be ideal, the reality is that data errors are normal. This year\u2019s Black Friday spend numbers<\/a>, for example, were under-reported throughout the day. There have also been multiple instances in which reported data is either inaccurate or incomplete, only to be corrected a few days later.<\/p>\n\n\n\n

While you’ll never control all elements of data quality, you can build systems to proactively and quickly detect problems, notify stakeholders, and take steps to address the issue. At Intentwise, we have multiple contingency plans based on the type of outage and the stakeholders needed to solve the issue. If you are building your own data solution in-house, plan for the resources and processes you\u2019ll need to handle the inevitable data errors.<\/p>\n\n\n\n

Infrastructure can slow as data grows.<\/h2>\n\n\n\n

Here at Intentwise, we believe there\u2019s no such thing as too much data. But if your data infrastructure isn\u2019t set up to scale and handle more and more data, you could run into issues like slow or non-comprehensive dashboards down the road, which could be costly. As you scale, you can maintain fast and accessible data by:<\/p>\n\n\n\n