How do you calculate true inventory carrying costs?

Add up your annual storage fees, insurance, spoilage and obsolescence write-offs, financing costs (if you're using credit to fund inventory), and the opportunity cost of capital tied up in stock (this can make or break early stage brands, particularly in an environment where investment capital is increasingly hard to come by). Divide that total by your average inventory value. Most CPG brands land between 20% and 30%. If you're higher, that's where to start digging.

What is a healthy inventory turnover ratio for a $10M CPG brand?

It varies by category, but most CPG brands should aim for an inventory turnover ratio between 4 and 8 times per year, meaning your average inventory sells through roughly every 6 weeks to 3 months. Perishable categories (refrigerated, fresh) should turn faster. Shelf-stable brands with long lead times might sit on the higher end. The key benchmark is whether your turns are improving year over year.

How does 'ghost inventory' impact my balance sheet?

Ghost inventory is stock that shows as available in your system but isn't actually sellable. It can be damaged goods, products that were miscounted, or items lost in a 3PL. It inflates your inventory asset on the balance sheet, understates your actual shrinkage costs, and causes you to under-order products you actually need. Regular physical cycle counts and 3PL reconciliations catch this early.

When should a brand hire a fractional COO for inventory control?

When inventory decisions are costing you more than the hire would. That usually shows up as recurring stockouts during promotional periods, out-of-control carrying costs, retail chargebacks you can't get ahead of, or a founder spending too many hours a week on supply chain problems instead of growth. A fractional COO or embedded ops team gives you senior-level oversight without the full-time executive cost.

How often should we conduct an ABC inventory analysis?

Quarterly is a solid cadence for most brands. If you're launching new SKUs or entering new retail channels, do it more frequently. Your 'A' movers change as your business evolves, and your safety stock and reorder points should reflect current velocity, not assumptions you made 18 months ago.

What is the difference between JIT and Safety Stock models?

Just-in-Time (JIT) means ordering inventory as close to when you need it as possible, minimizing on-hand stock. Safety stock is a buffer you hold above your expected demand to absorb forecast errors or supply delays. Most CPG brands need some version of both: lean on fast-moving, predictable SKUs, with a deliberate safety stock buffer on items where a stockout would be catastrophic (your top retail SKU, for example).

How do MOQs (Minimum Order Quantities) affect working capital?

MOQs force you to buy more than you might need in the short term, which ties up cash. If your co-manufacturer requires a 5,000-unit minimum but you're only turning 1,500 units a month, you're carrying roughly 3.3 months of stock from day one. Renegotiating MOQs is one of the most underutilized levers in CPG operations. Even modest reductions in MOQs can meaningfully reduce the cash you need to keep the supply chain running.

[FAQ QUESTION 1 — exact text from page]

[ANSWER TEXT — plain text, no HTML]

AI in Supply Chain - The Best AI Use Cases We’ve Seen in CPG Operations

Thought Leadership

Jul 31

Written By Josh Leider

You've probably had the following meeting at least 10 times this year - Someone pitches you a tool that promises to automate your demand planning, tighten your margins, and free up your team, all in one clean, killer dashboard. If you're running a consumer brand, the promise of AI in supply chain and ops sounds great, and that pitch really lands because your operations are the part of the business held together with spreadsheets, a boat load of data, late-night emails, and the good judgment of two or three people who have a majority of the process living in their head.

So the question isn't whether AI can help. It can, and if you're not using it somewhere in your operation you're likely leaving time and margin on the table. The question is WHERE it helps, where it quietly makes things worse, and what you should never ever hand it without a human auditing the work.

I've watched our team use AI tools across many many brands over the past couple of years, and I’ve started to pinpoint some consistent patterns.

Hand it the grunt work at the front of an engagement

The single biggest win is so boring that no one really talks about it. When you take on a new brand, or when your own team inherits a messy corner of the business, the first job is turning chaos into something you can act on. A pile of PDFs, scattered slack threads, emails with attachments nobody labeled, three versions of the same spreadsheet. Somebody has to read all of it and pull out data that matters.

That used to eat a huge amount of time. Now one of our guys can feed a whole range of materials Claude, get a clean summary, and know exactly where to point their attention before they start a real audit. On a recent client, that meant pulling numbers and analysis out of a dozen different sources, PDFs included, and landing it all in spreadsheets with almost no manual entry up front. It's not flashy, but it's probably the biggest efficiency gain we've gotten out of AI thus far.

If you do one thing with AI this quarter, make it this: Point it at the unglamorous data cleanup behind demand forecasting and planning that's currently stealing days from your ops team.

Build "what if" tools instead of guessing

The other place I’ve seen AI payoff is scenario work. Say you're launching a product and a component's running late. What does a two-week delay do to your on-shelf date? Or you're transitioning to new artwork and need to figure out how to manage the writeoff on the old packaging. These used to be half-day modeling exercises. Now you can stand up a working "what if" tool in an afternoon and actually pressure-test the tradeoffs before you commit.

One of our Directors of Ops built a SKU rationalization tool, essentially a SKU productivity engine, for a growing CPG brand, ranking every SKU against contribution profit, velocity, margin, and the cash tied up in inventory. What started as a one-off analysis became a live tool the founders run themselves, with export and margin views built in. But I’ll lean in on what actually created value here: it wasn't the AI doing the work. It was how fast it allowed us to iterate as the brand's questions changed. And none of it would have worked without our Director of Ops’ own years of experience deciding what to classify and how to calculate it. The client liked it enough that they showed it to a well-known SaaS product they were evaluating, and the vendor asked for a meeting to see how it was built.

The whole lesson in one sentence - the AI made an experienced operator faster, but the thinking that made the tool worth anything was still his.

Ready to stabilize ops and unlock growth?

Bravo CPG is the #1 fractional operations firm for growing CPG brands.

Learn More ->

Let it do first-pass document review, then verify everything

AI contract review is a natural next step here. Contracts and vendor agreements (co-mans, 3PLs, etc) are a natural fit. A first pass across a stack of agreements gets you to the risk clauses and open questions fast, instead of burning hours just finding where to look. Same with document review generally, where it's good at flagging language that could bite you later.

The word that matters there is "first." One of our COOs put it bluntly: triple-check the output. AI will make confident guesses in the gaps, and if you don't have the experience to catch a wrong assumption, you'll ship it. We had a client where someone used AI to draft a whole new supply planning model, and it looked fine on the surface. But leadership couldn't tell if it actually held up. So we audited what the machine built and flagged the logic that would have likely hindered the business. That's increasingly the job, closing the gap between what the tool produces and what actually survives contact with the real world.

Keep it away from judgment and relationships

Here's where I recommend holding the line (at least for now). This is where AI inventory management and demand forecasting still fall short, because demand planning and inventory management carry too much subjectivity to hand off cleanly. You can automate pieces of the repetitive work underneath them, and you should, but the modeling still needs someone who understands the business making the calls.

The bigger no-go is anything built on relationships. Negotiating a co-manufacturer agreement, coordinating a messy 3PL transition across multiple partners, managing a retail launch. That work runs on reading people, knowing when to push and when to hold, and adjusting in real time when something goes sideways. AI can prep the ground and organize the pieces, but it can't run the negotiation or own the relationship, and it can't be accountable when a call goes wrong. Those things get built over years, and they're nowhere close to being replicated.

Where the hype and the reality split

AI is still overhyped as a replacement for an entire ops team or for judgement based work. It's underrated at the front end of the work, the structured data grind that used to be pure time-sink. And it is not a silver bullet that thins out your team. The more useful framing is that a well-trained team, one that actually knows how to use these tools, gets meaningfully faster and sharper without the tools ever setting the direction on their own.

If you're staffing an operation for the next two years, the shift isn't fewer humans. It's humans pointed at oversight and judgment instead of manual assembly. Someone still has to check that the AI-built model holds up, the same way you'd audit a co-man's process before you trust it. The tactical work compresses. Accountability for getting it right lands on the same senior people it always did.

The brands that pull ahead here will be the ones whose people can tell, case by case, what to trust the machine with and what to keep in human hands. Judgment is the whole game.

Need Fractional Ops Support or Help Setting up AI in Your Supply Chain?

At Bravo CPG, this is the work we do every day. We're an embedded operations team for growth-stage food, beverage, beauty, and wellness brands, combining hands-on execution with senior-level ownership. We take real responsibility for production, co-man and 3PL management, demand planning, wholesale orders, freight, and everything in between. AI has made parts of that faster, and we lean on it hard where it earns its place, but the reason clients bring us in is the judgment behind the tools. Our goal is straightforward: help you scale profitably without operational chaos.

Josh Leider