In my last post I talked about how various supply woes, including inventory blindness, are a drag on performance marketing.
In terms of inventory blindness, the crux of my argument was that it causes you to pay for traffic that is extremely unlikely to convert.
In this post I’d like to discuss the steps you can take to improve inventory visibility to the great benefit of your performance marketing initiatives.
Four Steps to Improved Inventory Visibility
Here are four steps for improved inventory visibility and their application to performance marketing:
1. Static Forecast
Start by trying to achieve nothing more than the projected inventory around a product for the next 30 days. This can be enough to start moving your budget around.
Experiment first with skus in the bottom 10% of availability. Take your assortment that has the lowest 10% of availability, and look at your ad efficiency (ad spend / gross margin) on those products as compared with the rest of your campaign.
You’re very likely to notice large inefficiencies. If that’s the case then you can redirect that money elsewhere . . . theoretically toward items with high demand and high availability (though bid price dynamics obviously need to be factored in).
2. Batch status
The next step is being able to have a snapshot of inventory on a daily basis. This will allow you to apply a bottom 10-20% reallocation that should start to yield some very healthy returns.
When your analytics team runs the correlation between availability and revenue per click (RPC), they are going to list it as a top five predictor of revenue on acquired traffic. Not too surprising, because traffic coming through these channels is very ready to buy. After all, if you don’t have the product, your cross-sell won’t cut it. They’ll bounce and buy with someone who has availability.
Accuracy will be a challenge here, since inventory accuracy is only 55-65% across the industry. Luckily, there are ways to offset that by using a data sharing platform that can make sense of orders and inventory across every source. This is the idea behind a lot of Dsco’s data architecture.
If you have low confidence about the accuracy of the data, add a safety stock buffer–meaning when your stock appears to hit a certain number, pause spend and move it to a different product (you can read more on some of our anecdotal research about safety stock buffers in our next post).
When you’re able to take the next step by integrating the entirety of your organization’s inventory (internal and external) and feed it out in real-time via API, then you’re in a cool spot. Real-time matters most, however, if you can immediately react to what the data is telling you.
Here’s a simplified example of the value of real-time visibility:
Treestands are expensive, and buying traffic for them gets more expensive every day leading up to the opening day of hunting season.
- Assume you’re a week away from the opener, a new treestand sku has just come online and is being drop shipped by a supply partner (not surprising with how much space they consume). You’ve set your bids to take top placement for the first three weeks of the season.
- Six days away and you are running 140% efficiency (spend / gross margin) to strategically acquire these customers because of a belief that they’ll be coming back throughout the season.
- Five days away and by mid morning your hourly API call goes out to your supply partner.
- The delta between this inventory update and the previous update is massive (Dsco’s platform would see this and speculate a market purchase . . . meaning someone just bought the stands in bulk out of your supplier’s warehouse).
- You’re now out of inventory. Your ecom system processes the out of stock, but if you don’t get that ad spend reallocated immediately, you’re going to go from 140% efficiency to well over 500%, meaning that you will burn $5 for every dollar you bring in.
Unless a) this is part of a strategy to open a new, high growth market for you, like China, and b) you have a lot of cash to flow that spend, this is a very bad thing. By having real-time visibility to such changes via your data sharing platform, you’re able to pivot that spend immediately to more intelligent spaces in the market.
3. Momentum Intelligence
While many organizations are currently working toward achieving step three, it’s important to know that the next step beyond real-time reporting is even more powerful. This is the step of momentum intelligence.
Performance marketing channels are extremely bottom-of-funnel in most use cases, and it is common for these channels to run an auction model that drives first purchase margin to zero and beyond zero . . . classic prisoners’ dilemma. Momentum intelligence is when your organization is not just showing real-time inventory, but also looks at channel pacing toward sell out of that product.
Once you have this capability, you are enabling yourself to opt-out of the auction in smart ways. Here’s an example:
- You are a retailer with a strong presence in the kitchen category, and your summer assortment of mixers is now available online, in your stores, and in your fulfillment center with promised backup from your suppliers via drop ship.
- The assortment begins to sell down, and every order gets tagged by channel (having a good multi-touch attribution model is important here).
- Let’s assume that you bucket your performance channels into 3 buckets (Hot, Warm, Cool)
- Hot channels being those which have poor marginal spend efficiency, but potentially excellent volume (Google Adwords, etc)
- Warm channels being those with moderate spend efficiency, and reasonable volume (affiliate, social retargeting, etc)
- Cool channels being those with very healthy spend efficiency (SEO for example)
- By tagging sell-down momentum to each channel for a product, you now see that your SEO traffic will alone be sufficient to sell out all of these seasonal mixers
- You can therefore move your spend to pockets of the assortment that have a poorer sell-down momentum. You’re able to shift Adwords spend to other categories and reduce commissions to affiliates on these products in particular.
This concept doesn’t just apply for seasonal offerings, it works for any product where demand is high and restock may have some latency. Where this is happening, having good supply partners who are 1) telling you restock timelines systematically and 2) are willing to drop ship from factories to bypass the full truck route, will make these sellout momentum playbooks smarter.
A 12 month plan to move through these stages is more than aggressive but the sooner your organization creates high fidelity inventory visibility, the sooner the acquisition marketers in your organization can start playing the game by a whole new set of rules.
Your GMV and EBITDA will thank you.