Cost Optimization Isn’t Sexy Until You See What It Does To Your Bottom Line


I have previously shared the process that businesses can use to scale media and revenues.  If you’re a small company, that’s where the focus should be, because if you don’t have sales, the rest doesn’t really matter.

As you scale, though, it becomes crucial to make cost optimization an additional area of focus.  You of course want to get as much margin on those sales as possible.

I’ll acknowledge that optimizing your media, CPA’s, and revenues can be more fun and sexy than, say, lowering your costs to your shipping carriers.  Or working to improve your merchant processing fees. But if your larger goal is to put more money in your bank account or to have more money to invest back in the business, then shifting from what’s more fun to what is more impactful is an important change to make.

Below is a breakdown of 8 cost-focused categories where I’ve seen significant impact to a business’ bottom line.  As with any area that you want to improve, make sure there is a single person who is accountable for that part of your business.

  1. Merchant Processing
  2. Commissions to affiliates
  3. Commissions to sales reps
  4. Shipping / Fulfillment
  5. Technology
  6. Product Costs
  7. Customer Service


At the risk of repeating something from my prior article, I cannot stress enough how important prioritization is. Whether across cost buckets or in evaluating tests within one.  Sometimes you know straight away where your numbers are out of whack.  Even then, I’d suggest taking the 5-10 minutes to do some math on how much impact you think you can have.  Or how much improvement you need to get relative to other areas you may focus on.


A quick word on testing.  Generally, many of the areas below are more about vendor pricing and so it’s easy to assume that there really isn’t testing involved.  Most of the time, I’d agree with you.  But to the extent you are affecting customer experience, such as changing shipping carriers or repackaging your product, keep a particular eye out for noise from your customers.  These types of changes may not be based on a pure A/B test.  If you approach them as tests, however, then you and your team are likely to be more mindful of watching for changes elsewhere.

Let’s dig in…

1. Merchant Processing

This is arguably the least sexy of anything on this list.  It’s likely not the largest expense on your P&L, but it may be the one you cannot survive without.

If you can’t process payments, you are likely not in business.  And there are a TON of companies who are seeing their processor go away as a result of Operation Chokepoint, the government’s way to shut down marketers with questionable or problematic business practices.

When it comes to payment processing, many businesses optimize for ease and speed at the outset.  As such, they set up with Paypal, Stripe, or a 3rd party that works with smaller, oftentimes higher risk marketers.

But once you achieve any sort of scale, you may be leaving dollars on the table in higher processing fees because you have a sole provider or simply because you haven’t shopped around for a better partner.  Whether Paypal adds incremental orders for you is something you should test, but typically, going through one of these services means you have a higher overall rate or per-transaction fee.  Or how much you’re charged for chargebacks.  Or countless other buckets in that annoying bill that you aren’t really paying attention to.  That is real money out of your bank account.

You might also be eligible for better service, such as a lower reserve (or none at all) or getting approved to process more when your volume increase. Unfortunately, I’ve worked with several clients who weren’t able to scale, not because they didn’t have the demand, but because they literally couldn’t transact those orders.  Their merchant processing partner hadn’t approved them for higher volumes.  That is a painful, frustrating and costly situation in which to be.

Here’s the thing – merchant processing is one of the most annoying parts of running your business. And it’s an area that very few people understand.  But it could be costing you percentage points to your bottom line.  And those points are 100% margin, 100% cash.

So what to do? At the very least you (or someone on your team) should go through your statements to try to understand the different components of your bill.  There are consultants who can help you do this as well; it’s your choice whether you want this knowledge in-house or just want to pay for outside help.  Then get on the phone with your processor to talk through your bill.  Make it a goal to get better pricing and service for your company.

(I’m actually trying to get a friend to put out a marketer-friendly, no sales pitch webinar explaining some of the more important aspects of merchant processing. If that’s something you’d be interested in, shoot me a line here.)

So while merchant processing may not be as sexy as front-end split testing, when all is said and done, what margin do you get on that test that got you a 5% lift, versus the lower points you pay and better relationship you have with your processor?

2. Commissions to Affiliates

These next two sections arguably should’ve been in my prior article, as they are ultimately about media, but a) I forgot to put them in there; and b) they have some commonalities and relate to broader vendor and employee management, so they’re going here.

When it comes to affiliate commissions, I’m actually going to push back on a commonly-held philosophy: trying to pay out out as little as possible.  Many marketers see their affiliate channel, whether offline or online, as a hedge against paid media.  So they use it to balance out their numbers, in case their paid media comes in over goal.  I understand that mentality.  We did this for some time at Beachbody.  But then we realized that we need to manage our paid vs. affiliate traffic channels separately.  The latter wasn’t there simply in case the former had an off week so that our overall numbers looked okay.  Each had to perform on its own.   And our goal was to maximize the opportunity from each.

To do so, we, similar to most people who use affiliate traffic at scale, knew that we wanted to do everything we could to reduce friction for your affiliates – deliver to them ready-made banners, creative, swipe copy, etc.  And in that “etc.” bucket needed to be “give ourselves the best chance of maintaining that partner’s traffic by making working with us as lucrative as possible..”

Most affiliates are driven predominantly by how much money they make.  That’s their business, they’re allowed to think that way.  So while you might be hedging against your cash media, it’s also likely that you are not getting the traffic you could.

This isn’t to say you are blind to issues like fraud or attribution.   Not to mention remembering to nurture the relationship, since behind every affiliate name is a person.  But I’d also suggest looking at the value of customers coming through you affiliate channels.  Figure out what margin you’d be happy with.  There isn’t a great reason why that margin should be any different as that from your paid media channel.  Then payout as much as possible.

Once you know that figure, you can test out different payment structures – how much is paid day 1, over time, or even whether you extend lifetime commissions to your partners.

3. Commissions to Sales Reps

I could make the same argument here re: payout amounts as for affiliate commissions above, so I’ll add an additional nuance.

And it starts with a small point I made just above.

Remember:  Your sales reps are people. 

As such, each is motivated by different things.  What you have to do is figure out a model that works for your business and then attract those for whom that model works.

One area where I’ve seen people make huge gains with their salespeople is in creating the right incentives for the behavior and results they wanted to see.

On one side, paying your salespeople too high of a base salary might result in their doing only as much to not get fired.  (You already have this issue with most employees.)  On the other side, if you pay 100% commission, that might mean too much financial stress for your folks.  In fact, I’ve seen businesses put their sales reps on $500 per week for the first 3 months just so they have some income coming in.  But it’s not so unwieldly that it’s a massive risk for their org.  And then I’ve seen these same companies have some high 6-figure and 7-figure earners from their sales department.

To be clear, there is no right answer.  I have some friends who will argue for hours that any commission is a bad thing.  That the best way to get people to deliver their best is by paying only a salary.  Others see that as the worst thing you could do.  You need to find the model that works for you.

My goal here is simply to raise some considerations to help you think through different structures.

As if there was any doubt, sales is a brutally hard job.  More than 50%, possibly 90%, of the time, you’ve just got the wrong person in that role. And it has nothing to do with comp structure.  There is only so much you can do with the comp structure if you don’t have the right people in the right role.

But if you feel like you do, figure out how to craft a model that works for them and the business.

4. Technology

I’m intentionally steering clear about specific tools or technology partners.  Those choices are too business-specific.  Instead, there are a few processes I’ve seen have great results regardless of business type, technology partner, or otherwise.

The Ask

Last month I was a conference where one of my friends negotiated $7K in monthly savings from 2 different partners – his ESP (email services provider) and his ecommerce platform partner.  Better yet, he got those reductions simply by asking.

Whether or not $7K per monthly is a rounding error for your business or a significant part of your take-home, that’s $84K in real savings.  And this was a reasonably low effort way to reduce costs.

My point here is in some ways the same as elsewhere – you should take a regular review of your partner fees.  And sometimes it’s just as simple as asking for a reduction.


This is going to sound so obvious, but I can’t tell you how many times I’ve seen it make it a big difference.

On a regular (at least once a year), audit of your technology expenses, particularly those with a monthly plan on a credit card.  It’s likely you signed up for some service (or seven) that was a trial that rolled to full price (you’re familiar with that model, right?).  Or someone who is no longer around put one in place.  Regardless, if no one is leveraging that tool, it’s a waste of money.

Some of these services may be $50 per month, some may be several thousand per month.  For starters, if you’re not using something, cancel it.  Just as important, doing this type of review is a healthy way to reinforce to the rest of your organization that you don’t tolerate waste.

(I’d also suggest doing this with your personal credit card.  I’m guessing you have a bunch of similar types of services that you continue to pay for but haven’t gotten any use in months.)


This is by far the most difficult of the three, from almost every possible perspective.  Ultimately, this is about looking back on where you’ve made technology (or frankly other) investments.  Ideally, you evaluate ROI, or at the least, examine how the final costs of certain projects and what you got out of them, even if qualitatively.

Whether in a startup or in a billion-dollar plus organization, people complain about IT.  Tools don’t do what people want. Everything takes too long.  Projects go over budget.

So why does that the adage of “what gets measured gets improved” not seem to apply to IT?  If you’re a performance marketer, then measurement is a hallmark of your business.  And yet so often, whether quantitatively or qualitatively, rarely do we conduct a look-back on IT projects to see what went well (or didn’t) and what, particularly process-wise, can be done better the next time around.

At the least every quarter, sit down with your teams to conduct a formal review conversation.  Leave ego, blame and excuses at the door (I know, easier said than done).  And have some honest and service-driven conversations about how you can maximize your investments in IT, arguably one of the more important enablers (or inhibitors) of your business.

5. and 6. Shipping / Fulfillment

If you don’t ship a physical product to your customers, then by all means jump to the next section.  But I’m going to lump both of these items together here.

A few questions to start.

If you do ship a product, do you know:

  • Do you know your average per order shipping cost?
  • Assuming you ship different products or different configurations, do you know the cost by these subsets?
  • Given that shipping rates are grouped above certain weights, do you know your tiers and have you looked to see if there is a way to decrease the weight of your higher-volume packages?
  • If you ship product to Alaska, Hawaii, Canada or anywhere else internationally, have you examined shipping costs and/or customs and duties fees. It’s possible that customers in these regions are not in fact profitable to the business.

With your fulfillment partners, do you know:

  • The average cost of processing your product through your warehouse partner is?
  • If you have multiple items in an “typical” order, have you explored whether it would be cheaper for your warehouse to pick and pack each order real-time versus “pre-kitting” those orders beforehand.
  • What about storage costs? I’ve got a couple products that I keep at the freight-forwarder at the Port of Los Angeles because their storage fees are so much cheaper than the warehouse’s (or Amazon’s).  I didn’t think storage fees were that big a deal until I got an unwelcome buzz-kill from an invoice late last year.

Or how about something super simple – what percent of your net revenues are shipping and fulfillment?

I’ve got some folks in the physical products space where shipping is ~7% and fulfillment is ~3% of net revenues.  This is such a bad number to compare to because there is no context of their business.  But do you know your numbers?

Getting a benchmark is always the first step towards improvement (“what gets measured gets improved”, right?).  Then, just like all the other parts of your business, it’s a matter of identifying leverage points and working to chip away at them.  Whether it’s changing the way you all operate internally or getting more out of your vendors, you have to put some focus towards these areas.

Even then, sometimes your partners don’t respond.

One of my recent clients had gotten so frustrated with their fulfillment partner that they began in-depth conversations with several alternative vendors.  When they came back to their current partner to talk through why they were seriously considering switching vendors, what became clear was their 3PL’s pricing was actually pretty decent but that their service sucked.  And that was in large part because their left wasn’t talking to the right.  I’ll admit that typically a partner will jump to fix things and then service will fall off again in a short while.  In this client’s case, their current partner, to their credit, has raised and maintained their level of service.  Which is a win for both parties.

Many times, unfortunately, a change is in fact necessary.  And as much as it takes real work to switch vendors, when you know it’s the right thing to do, it’s then a matter of getting it done.  (The same can be said with any relationship.)

Yes, it’s frustrating to have to get to this point, but it all begins with someone having accountability for this part of the business, and then making sure that pricing, service, and anything else you deem important is at the level you demand.  And deserve.

A few other notes here.

Service Levels

Amazon has changed the game in terms of expectations on shipping.  And yet, if you look at a lot of offers running on TV, they rarely tout next-day shipping or even 2-day shipping (which is considered rush).  They can get away with several business days to get their product to their customers.  In part this could be explained because the typical buyer of a TV-marketed product is a bit older, but that can’t explain everything.

Sure, these marketers are likely losing some customers in not offering Amazon-type delivery.  But not everyone is Amazon.  There could be financial, logistics, or technology reasons that limit your ability to get your product to a customer next day.

And you know what? That might be okay.  At least for now.  Because going with a service like Fedex Smart Post, which can take 3-6 days, might be more economical for you.  Even when you factor in the impact on conversion, additional contacts to your customer service department, or to your brand.  But you can’t always optimize for everything.  And at the least, you should be intentional about what you’re doing and know your options.

Direct vs. Through a Partner?

At what point does it make sense to try to negotiate your shipping costs directly with a carrier as opposed to through your 3PL partner?  Many folks manage their shipping agreements through their warehouse partner. At the outset, this is probably a good idea. As your volumes increase, it might make sense to explore what you can negotiate directly.  (I’m assuming that you’re shipping outside of Amazon.)

As a somewhat arbitrary guide, I’d say once you are shipping several thousand units per week, it’s time to have these conversations.

What you may find is that your 3PL, because of the sheer volume they process, gets better rates than you can get.  But it’s possible that:

  • You find that they haven’t negotiated their terms in a while.
  • You might learn about certain fees or services that you were previously unaware of
  • Building the relationship directly might allow you to resolve certain issues in the future, if only because you know the vendor directly. During my time at Beachbody, I cannot stress how important these direct relationships were, even when we continued to work through vendors or agencies.  Simply being able to pick up the phone to avoid any game of “operator” saved us a ton of money, not to mention time.

The worst that can happen is the carrier says your volume isn’t high enough or that your rates don’t change.  But hopefully at the very least, you’ve started to build the foundation for a direct relationship.

One Warehouse or Multiple Locations?

Here are some of the considerations to answer that question.

  • First you need to know the current geographic distribution of your customers.
  • Once you have that, then it’s a matter of understanding shipping cost by zone (how far away from your fulfillment center you ship to).
  • Does your current 3PL actually have multiple locations?
  • What is the net additional inventory position you’ll have to have in place, and what is the financial exposure of that additional inventory? (It’s never a matter of splitting up your existing inventory evenly into the new warehouses. And since inventory = working capital = cash, that isn’t a no-brainer decision for anybody.)

Multiple warehouses may not be applicable for all businesses, but it might be worth looking into.  It might mean net additional inventory and some additional costs with your current 3PL.  It might mean exploring a new partner.  It also might mean lower shipping expenses and few customer service inquiries.

7. Product Costs

Beyond the “ask for a discount” point I’ve made arguably too many times, let’s talk about China.

Conventional wisdom used to be that manufacturing a product in China is cheaper than producing it in the US.  I’d suggest that in many cases that is correct.  But I’d also suggest that it doesn’t capture the fully-loaded cost of delivering a product to your warehouse.

In particular, it misses the cost of shipping your product across that big ocean over there.  If you were one of those unfortunate folks who got caught out when the Port of LA workers went on strike last year, it also misses the cost of being out of stock.  I know of too many 8- and 9-figure businesses that lost millions in sales because their product was stuck in a container.  They literally couldn’t access it even though it was in port.

In addition, when you manufacture in China, you have to consider a couple other items: 1) a month off for Chinese New Year (sure, it’s not a month-long holiday but the effects are at least that long); and 2) shipping via boat is a lot cheaper than via air.  But it’s also a lot slower, so you’ve got a time-in-transit issues to content with.

Finally, know your customer.  In some markets, being able to say “Made in the USA” can be an important marketing tactic.  In others, no one is paying attention.  Or maybe no one has suggested that the customer pays attention and it can turn into a competitive advantage for you.

If there’s one theme that has resonated throughout here, it’s to make sure someone has done the diligence to explore alternatives.

8. Customer Service

Let me start with here with a sampling of metrics with which to manage customer service:

  • Customer Satisfaction (CSAT) / Net Promoter Score
  • Contacts – phone, email, chat
  • Total costs
  • Cost per contact, cost per initial order
  • Calls offered / Calls Handled / Abandonment rate
  • Speed of Answer / Avg. Handle Time / Avg Talk Time
  • Calls answered within SLA’s
  • Disposition Reports – reasons for calling, reasons for cancelling
    • Wizmo (“Where’s my order”), refund request, continuity cancel, escalation, etc.
  • Cancel-Saves
  • AG / BBB complaints
  • Revenues / Margin generated and vs. costs
  • Reship rate

That’s a lot there.  But I find that customer service is one of those areas that needs a particular form of structured metrics to best manage it.  Not to say that the numbers are the only way to manage customer service.  But when you look at that list above, you realize just how much specific information you can have about the performance of your call center.  And that makes identifying areas of improvement that much easier.

Sales vs. Care

Perhaps one of the most important points I can make about customer service is that reps who are great at customer care are not the same ones who are great at sales.  Sure, there may be those edge-case examples of reps that can do both great.  But it’s crucial to approach the call centers differently.

Everyone should be customer-focused.  Everyone should have a sensibility for brand.  Everyone should be aligned with the company values.

But make no mistake.  A sales floor and a customer service department are fundamentally different environments, with unique cultures, and frankly with entirely contrasting personalities. 

So much so that you may want to change how you funnel certain types of customer service calls.  For example, I’ve tested sending continuity cancel calls to sales reps instead of customer service reps.  And found that in most cases, the sales reps do better.  Even after factoring for follow-up calls, discounts, and refunds.

At the same time, if you put a sales person on a customer service floor, 99 times out of a 100 they will not succeed.  This isn’t a criticism, nor a character flaw.  Plain and simple, it’s about putting people in the right roles.

Final Thoughts

I’ll go back to what I said at the beginning.  Without sales, cost optimization is generally moot.  But once you’ve achieved a certain level of scale, your business demands it.

And as the business grows, presumably you are bringing in specialists to focus on various parts of the business.  Task them to set benchmarks and then to work to improve their business area’s performance.   Just as you would for the seemingly sexier parts of your business, like media and revenues.

When I read the book “The Millionaire Next Door,” I remember one big lesson loud and clear.  That there are two components to a P&L.  Revenues. And Costs.  Ignoring the latter can make all the front-end work you do go for naught.  Not to mention, managing the latter can mean less pressure on the front-end.

And managing both revenues and costs?

That’s where scale really happens.



Get My Proven Process to Scale Your eCommerce Business AND Build Your Brand

See the exact analyses that I conduct for my clients & the actions to prioritize based on those learnings. I also include a section specifically for subscription businesses and a bonus on crafting incredible customer experiences.

Featured Articles

Babak’s approach is equal parts brand-based and data-driven analytics. He brings deep practical experience to help frame each decision we make along the way.

Caith Chapman
GM, eCommerce & Consumer Products, Tastemade

Reader Interactions

Leave a Reply

Your email address will not be published. Required fields are marked *