The strength of your brand identity can be the edge you need over your competitors. You can’t beat every single competitor on price alone—but price isn’t everything when your brand stands for quality and innovation.
Whether you’re selling on Amazon, Walmart, or your DTC site, you want to ensure you are crafting SEO-optimized pages that also help elevate your brand.
If you created a new sub-industry, tell the story of why. If you’re in a crowded field, prove that you’re different.
In this webinar, Emma Schermer Tamir, CEO of Marketing by Emma, will show brands how to craft compelling listings and content that position them for success.
Learn how to:
Register now, and join us April 25 at 10 am PST / 1 pm EST.
.jpg)
For years, a debate has raged in the industry about how to manage bids on Amazon. Do you opt for a rules-based approach to change your bids—essentially, following a series of automated “If X, then Y” statements—or do you use an AI?
We have heard all of the arguments on either side, and ultimately, in our Ad Optimizer, we have chosen a hybrid approach: we allow for brands to mix-and-match between rules-based and AI-based bid management. That’s because we think there are benefits and downsides to each approach.
If you’ve been in this business for long, you’ve probably been pitched an AI-based bidding tool. But how do these AI systems work? How are they trained, how do you monitor their success, and what are the pros and cons as compared to rules-based bid tools?
In the following webinar, Ryan Burgess, Intentwise’s head of growth, and Sreenath Reddy, Intentwise’s CEO, break it all down. They discuss:
At Intentwise, we take a set of historical ad data, and we split it up—80% of that data is used to train the model, while the remaining 20% is used to test the model. We continue to tweak it until our model spits out the expected result.
We train our AI models on metrics like impressions, clicks, and conversions as well as retail signals like product price and inventory. When you feed these inputs into the AI model, it begins to establish the relationships between each.
Let’s say you’re bidding up your ads on a keyword in order to get more impressions. If you reach your maximum ACOS, a purely rules-based model will tell you to bid down, to level out your ACOS. Sometimes that makes sense—but sometimes the solution to improving your ACOS is, counter-intuitively, to bid up. Bidding up might, for instance, increase your share of top-of-search placements, which in turn could drive exponentially more conversions.
Rules-based models can’t catch these nonlinear nuances; only AI can.
Conversion rates on ads don’t always perform how you expect. In the days leading up to a big sales event, like a Prime Day or a Black Friday, conversion rates have a tendency to plummet. That happens across the board, and is not specific to your campaign.
Caution is warranted. A rules-based model might tell you to bid down, but that can pose problems. When, say, Prime Day happens, if you’ve bid down too much, your ads will get buried. Only an AI will be able to correctly attribute a sudden conversion rate drop might to the time period, and hold your bid steady.
AI algorithms aren’t perfect on their own, and even the well-trained ones sometimes fail to hit the ACOS targets you set. You need a human to monitor these models, conduct quality control assessments, and continuously mix in new inputs to refine the results.
