SFIX by the numbers

[Smorgasbord1]

Stitch Fix just put up a presentation on how the business is doing: https://investors.stitchfix.com/static-files/ec8ba972-9fde-4… This post pulls numbers from there.

Let’s start with subscribers, I mean “Active Customers” (although to me a subscriber is more valuable due to the implicit recurring nature). Here’s the YOY growth:
2016-2017: 31%
2017-2018: 26%
2018-2019: 18%

SFIX’s Active Customer growth is steadily declining.

OK, but what are the actual growth numbers? Glad you asked:
2016-2017: 520K
2017-2018: 568K
2018-2019: 494K

SFIX added fewer new customers last year than in the prior two years.

Not a good sign, especially considering that advertising spend is increasing every year:
2016: $21.9M
2017: $68.4M
2018: $98.2M
2019: $157.8M

So, each new Active Customer in 2019 cost around 3X more than it did in 2016!

Gross Margin has been flat for 4 years now:
2016: 44%
2017: 44%
2018: 44%
2019: 45%

Profits. Maybe profits shouldn’t matter - I’d argue that at this early stage SFIX should be pouring almost all profit back into the business, but since being profitable is often cited as a reason to invest, let’s look at the EBITDA numbers:
2016: $73M
2017: $61M
2018: $54M
2019: $40M

Declining profits. Steadily declining. Maybe that should be viewed as a good thing?

But, Operating Profit has declined from 9% in 2016 to 1% last year. SFIX’s own target is 10%-12%. How are they going to get to that? Should they even be talking about that now?

Don’t get me wrong. This isn’t a failing business. And I really like the Data Science approach they use to make customers happy, to manage inventory levels, and to track how they business is doing. Good stuff. It’s why I was interested in the company and make a tiny investment.

But, I haven’t seen other important numbers; for instance, on client retention. Is the declining rate of customer acquisition (and increasing acquisition costs) reflective of customer burn-out or a difficulty in finding new customers (or some combination of the two)?

Revenue is growing, but not accelerating:
2016-2017: 34%
2017-2018: 26%
2018-2019: 29%

That’s probably simply a function customer growth in numbers being flat/declining while being able to get a few more dollars out of each existing customer each year. It seems clear that their approach means that customers who stick with the platform are more and more satisfied with it. That is, the longer a customer provides data to Stitch Fix, the better Stitch Fix can figure out what will make the customer happy. Data Science works!

But, I see customer acquisition as a hurdle for the company, and I feel this is borne out by both the increasing costs as well as indicated by the constant stream of new programs the company is rolling out. There’s a monetary risk for new customers - if they don’t like any of the 5 items in their “Fix” they’re out $20. I think of all the times I’ve walked into stores looking for something and didn’t buy anything - imagine if that cost me each time - and that’s a whole store with hundreds of items!

And if you read between the lines on what the company says, for instance from https://www.fool.com/earnings/call-transcripts/2019/06/05/st… :
There are predictive algorithms we believe we’re better able to reach clients who are a good fit for our service and who we are able to retain for a longer period…The first is we’re getting better at attracting quality clients. Katrina talked about our ability to better target clients who are right fit for our service. (emphasis added by me)

They’re acknowledging that their service isn’t right for most. So when in their presentation they show a TAM of $93B (online being 21.7% of the Apparel, Footwear, and Accessories Market), that’s disingenuous in my view. They already know and acknowledge their approach has limited appeal. The question is how limited? With competition from brick and mortar retailers like Nordstrom (Trunk Club) as well as online retailers like Amazon (Personal Shopper), Stitch Fix doesn’t appear to have a moat and so will survive on how well it competes with these pretty good companies.

If Stitch Fix really believed in their platform, they would roll out Style Shuffle to anyone, and when enough choices were made to build up a full customer profile, send a first Fix with no potential $20 fee. They don’t do that probably because it would encourage people who aren’t “the right fit” for their service to try it out, and so the success rate might not be so good, even with great Data Science. Or maybe they feel non-customers won’t spend enough time on Style Shuffle (but that would actually be a good thing because those who stick with it are probably more likely to buy!).

Penultimately, the number of new business models that Stitch Fix has rolled out recently is potentially both encouraging and discouraging. Style Pass moves the nothing per Fix fee out to a single yearly fee, so might be perceived by customers as less risky. But, the company isn’t rolling that out to everyone, perhaps due to potential costs of sending out multiple “nothing-liked” Fixes. Then there’s 3 new programs that are basically ways to present a limited set of choices to customers for them to buy directly: Extras, Shop New Colors, and Shop Your Looks. I appreciate the argument that Stitch Fix can leverage their platform to do a better job at presenting choices to customers, but whether that is really true is far from certain, and I’m not convinced that an Amazon or Banana Republic/Gap or Nordstrom couldn’t do the same. And that there’s so many of these new programs might be indicative of the company recognizing that Fixes doesn’t have legs.

To be fair, some of the data presented by SFIX is promising. Customers getting happier the longer they’re in the program (eg, 8% more US Women keeping at least one item in their Fix and looking forward to the next Fix). Retention rates for newly acquired customers being higher than the rates were a couple/few years ago. But, so far anyway, those improved internal metrics haven’t translated to improved business results - at least not the level I’d expect/want.

At the end of the day investing in Stitch Fix requires believing in Katrina Lake, her management team, and their data science approach to making customers happier at lower costs than the competition. Right now the Data Science on their results doesn’t give the same view as the Data Science on their customers’ happiness.

[SaulR80683]

Let’s start with subscribers, I mean “Active Customers” (although to me a subscriber is more valuable due to the implicit recurring nature). Here’s the YOY growth:
2016-2017: 31%
2017-2018: 26%
2018-2019: 18%

SFIX’s Active Customer growth is steadily declining.

OK, but what are the actual growth numbers? Glad you asked:
2016-2017: 520K
2017-2018: 568K
2018-2019: 494K

SFIX added fewer new customers last year than in the prior two years.

Not a good sign, especially considering that advertising spend is increasing every year:
2016: $21.9M
2017: $68.4M
2018: $98.2M
2019: $157.8M

So, each new Active Customer in 2019 cost around 3X more than it did in 2016!
…

Smorg, what an incredibly clear analysis. Thanks so much for your time and effort. And it’s actually even worse than you said: $157.2 divided by $21.9 equals 7.2, so with roughly the same number of new customers it’s costing 7.2 times as many dollars per new customer.

Saul

[imuafool]

Without full knowledge and understanding, simple math can be very misleading.

Believe it or not, there is a SFIX corporate strategy in place that reveals and explains why the growth of active customers has been declining and why advertising costs have increased, while year-over-year quarterly growth in net revenue has been trending upward. Back on 6/9/2019 in my SFIX Q3 post at this board, I related the following:
https://discussion.fool.com/stitch-fix-up-24-after-q3-results-34…

[My comments:
Below I show both the year-over-year and sequential quarterly growth of Active Clients.

**Period  Active Clients    Y-O-Y    Sequential**
 **(in millions)   % Change    % Change**

Q3 ’19     3.133          16.6%       5.8%
Q2 '19     2.961          18.1%       1.1%
Q1 '19     2.930          22.3%       6.9%
Q4 '18     2.742          25.0%       2.0%
Q3 ’18     2.688          29.6%       7.2%
Q2 ’18     2.508                      4.7%
Q1 ’18     2.396                      9.2%
Q4 ’17     2.194                      5.8%
Q3 ’17     2.074

While year-over-year quarterly growth in net revenue is trending upward, the table above shows y-o-y quarterly growth in active clients going the other way. Here’s what Stitch Fix management and performance have done.

• To reiterate, net revenue per active client has been rising for the recent past four consecutive quarters, largely driven by continued strength and improving retention in Women’s category, as well as success in ongoing strategic initiatives.

• As mentioned above, management is focusing on acquiring high-quality clients, which has driven stronger retention in Women’s. At a recent JP Morgan Global Technology, Media, and Communications Conference, Stitch Fix CEO Lake conveyed her company’s preference for finding high-value customers via targeted approaches, where the cost of acquiring a new client is more expensive, but the payback is quick and profitable. The company is directing its acquisition efforts on the kinds of shoppers who are most satisfied with Stitch Fix versus broad brush approaches that bring in a higher number of customers that may or may not be the best fit for the service. Also, CEO Lake related that when faced with higher than expected customer demand during Q2 2019, the company decided to defer advertising dollars to later in the year. Why? They strive to balance customer acquisition efforts to ensure that new clients are not brought in when inventory is tight that would reduce available selection and likely result in less satisfactory experiences for customers.

• Also mentioned above, the company delivered strong client and business outcomes through Style Pass. A year ago, Stitch Fix introduced its Style Pass - a $49 one-time charge that gets a client 12 months of free styling. As a bonus, $49 is added as credit to a client’s account that will be applied to the client’s next purchase(s). A client’s Style Pass membership will automatically renew after one year. Stitch Fix will charge $49 to the member’s card on file, and the member will receive another $49 in account credit. To cancel the renewal, the member emails hello@stitchfix.com before the member’s Style Pass renews.
At the 6/5/2019 earnings call, CEO Lake reported: As you know we are constantly looking for new ways to engage and delight our customers and Style Pass which just celebrated its first anniversary is a great example of it. Today our one-year renewal rate have exceeded 70% across both men’s and women’s clients. In addition, as of Q3 '19, Style Pass has continued to reduce friction from the client experience and deliver better client and business outcomes. Specifically, the program improved client retention and increased average revenue per client and client satisfaction as compared to non-Style Pass client. We will continue to roll-out Style Pass to clients in a disciplined manner to ensure that the program benefits both our clients and our business.]
—————————————-

Now in the latest Q4 and Fiscal Year 2019 Letter to Shareholders, Katrina Lake, Founder and CEO, Mike Smith, President and COO, and Paul Yee, CFO provided further clarification on their corporate strategy as follows on this matter:
https://investors.stitchfix.com/static-files/ef414eda-2498-4…

Supplemental Performance Disclosures:

We continued to generate fast payback on our marketing investments.

Our success in acquiring high-quality clients gives us greater confidence in our marketing strategy and our ability to drive fast paybacks. One way we measure payback is by analyzing cumulative gross profit generated by a given cohort relative to advertising spent in the quarter we acquired those clients.

The chart below shows our quick payback and high return on advertising investment for four different cohorts and how this return increased for cohorts with longer tenures. Specifically, the chart reflects the cumulative gross profit generated by clients who checked out their first Fix between October 29, 2017 and January 27, 2018 (Q2’18 cohort), January 28, 2018 and April 28, 2018 (Q3’18 cohort), April 29, 2018 and July 28, 2018 (Q4’18 cohort), and July 29, 2018 and October 27, 2018 (Q1’19 cohort).

We determine the marketing payback ratio for a cohort by dividing their cumulative gross profit by the corresponding quarterly advertising spend. The chart below demonstrates how we continue to drive near-term payback on marketing spend and that the marketing payback ratio grows as clients engage with us over time.

• In Q1’19, we spent $39 million on advertising to attract clients who, as of Q4’19, have already generated $94 million of cumulative gross profit in the three quarters since they joined, reflecting a marketing payback ratio of 2.4x.
• In Q4’18, we spent $29 million on advertising to attract clients who have generated $80 million of cumulative gross profit in the four quarters since they joined, reflecting a marketing payback ratio of 2.8x.
• In Q3’18, we spent $25 million on advertising to attract clients who have generated $104 million of cumulative gross profit in the five quarters since they joined, reflecting a marketing payback ratio of 4.1x.
• In Q2’18, we spent $20 million on advertising to attract clients who have generated $85 million of cumulative gross profit in the six quarters since they joined, reflecting a marketing payback ratio of 4.3x.
————————————-

Please go to aforementioned website to see chart and all footnotes for a complete understanding of this SFIX strategy.

I’m currently in Hawaii with my family for 3 weeks, spending quality time with my 95-year old mother and will have no time to provide a comprehensive evaluation of SFIX Q4/FY 2019, except for the above response to a very misleading take in the OP.

Aloha & Regards,
Ray

[Smorgasbord1]

And it’s actually even worse than you said: $157.2 divided by $21.9 equals 7.2, so with roughly the same number of new customers it’s costing 7.2 times as many dollars per new customer.

My math was:
2017: $68.4M spent to acquire 520K new customers. That’s $131.54 per new customer.
2018: $157.8M spent to acquire 494K new customers. That’s $319.43 per new customer.

The presentation didn’t show how many new customers were acquired in 2016 over 2015. And it’s probably fair to say that some of the advertising was to get existing customers to spend more - but that’s probably a very small portion of actual spend since Stitch Fix has their existing customer direct contact information.

Still, the company is profitable, so clearly each customer is worth several hundreds of dollars to Stitch Fix, at least over time.

[Smorgasbord1]

I wrote:
2018: $157.8M spent to acquire 494K new customers. That’s $319.43 per new customer.

Ack! that should be 2019 of course.

[Smorgasbord1]

Without full knowledge and understanding, simple math can be very misleading.

Hi Ray - thanks for the counter view. After reading what you wrote, I fail to see where I was misleading anyone.

there is a SFIX corporate strategy in place that reveals and explains why the growth of active customers has been declining and why advertising costs have increased

OK, so first, we agree that growth of active customers has been declining and we agree that advertising cost costs have increased.

Second, I didn’t mislead anyone on why this is happening, I even quoted the conference call where they talked about finding not just any customers, but customers who are “the right fit for our service.” I also pointed out the 8% increase in happy US Women customers and the higher retention rates (but no numbers supplied for the latter).

Third, the non-quantitative statements about improved retention rates and increased revenue per client are fine, but the numbers aren’t given to us (that I found, maybe people who’ve been following the stock longer know more). More importantly, however, this hasn’t translated into better business results. Gross Margin is flat. Profit is down. Total revenue isn’t accelerating and isn’t growing significantly faster than customer acquisition, so my conclusion is that the increased revenue per client is only a smallish increase.

Also, CEO Lake related that when faced with higher than expected customer demand during Q2 2019, the company decided to defer advertising dollars to later in the year. Why? They strive to balance customer acquisition efforts to ensure that new clients are not brought in when inventory is tight that would reduce available selection and likely result in less satisfactory experiences for customers.

I can’t understand why being forced to slow down your company’s growth is somehow a good thing. Yeah, it’s better than losing your existing customers, but it means that if Gartner were rating SFIX they’d be in the bottom quadrants since their “Ability to Execute” is low. Whether the company needs more money or just needs to spend the money it has more wisely I don’t really know, but the thing we look for in choosing companies in which to invest is keeping existing customers happy AND bringing in new customers. Not either-or.

Also mentioned above, the company delivered strong client and business outcomes through Style Pass.

Yeah, I talked about this and a few other new initiatives as well. On the face of it, the success of Style Pass shows that the original “nothing liked” Fix $20 penalty was maybe not the best business model to adopt. But, even management doesn’t seem convinced Style Pass is all good as they still are rolling it out slowly to customers. I don’t believe a new customer can join and immediately use Style Pass, for instance. If this was so great for SFIX, wouldn’t management roll it out to everyone immediately? What’s the down side here - and why doesn’t management talk about it?

And, as I said, the other initiatives strike me as new direct buy alternatives. They’re not Fixes, they’re “simply” ways to present items to customers that they are more likely to buy. As I said, that’s great - I love to see Data Science be used this way. But, it hasn’t translated into better business results: customer growth and revenue growth are flat at best, while profit is consistently declining.

[XMFKBecks]

And, as I said, the other initiatives strike me as new direct buy alternatives. They’re not Fixes, they’re “simply” ways to present items to customers that they are more likely to buy. As I said, that’s great - I love to see Data Science be used this way. But, it hasn’t translated into better business results: customer growth and revenue growth are flat at best, while profit is consistently declining.

The direct buy feature (Shop Your Looks) is BRAND NEW. It is in Beta testing. It has been introduced to a portion of women’s clients only, and only for a short time. The company shared some VERY EARLY results.

Shop New Colors has been around for ONE QUARTER.

There is no expectation at this time for big business results YET. They are laying the groundwork here for the LONG TERM.

Karen

[captainccs]

A very good thread indeed!

I can’t understand why being forced to slow down your company’s growth is somehow a good thing. Yeah, it’s better than losing your existing customers, but it means that if Gartner were rating SFIX they’d be in the bottom quadrants since their “Ability to Execute” is low. Whether the company needs more money or just needs to spend the money it has more wisely I don’t really know, but the thing we look for in choosing companies in which to invest is keeping existing customers happy AND bringing in new customers. Not either-or.

Being forced to slow down your growth is not a good thing but it’s part of managing reality and to me it shows very credible cashflow management. I like it.

Also, CEO Lake related that when faced with higher than expected customer demand during Q2 2019, the company decided to defer advertising dollars to later in the year. Why? They strive to balance customer acquisition efforts to ensure that new clients are not brought in when inventory is tight that would reduce available selection and likely result in less satisfactory experiences for customers.

The alternative is debt or dilution.

Denny Schlesinger

[XMFKBecks]

As a Stitch Fix customer, the way I managed the $20 risk of getting a box (before I was offered Style Pass), is that I would throw a pair or two of underwear from the Extras program on my order. If it was an 0/5 box, I recouped most of my $20 on inexpensive product that I knew I would keep and use.

The Extras program offers underwear, bras, shapewear, pajamas, tights, camisoles (for women - I have not seen the men’s items), and includes brand-name as well as house brand (Everyday by Stitch Fix) items.

Karen

[XMFKBecks]

In regards to looking for customers who are “the right fit”, this is smart. They are looking for customers who will generate profits and spend money. They have worked to reduce/eliminate shrink (theft). Some customers are a PITA and don’t buy anything. They don’t want those kinds of customers. Some customers spend hundreds a month. They want those customers. It absolutely makes sense.

There are enough high-spending customers around and Stitch Fix wants to capture them. They should capture the best, most profitabble clients and let the cheapskates go to Amazon.

Karen

[BroadwayDan]

I consider myself both a Saulinian investor and Stitch Fix fan. I own a 4-5% position. But it raises a good question that I’ve discussed w my pal, Phoolio18.

The way I see it is the Saulinian method tries to get as close as possible to objective truth. Because A and B are actually happening it will lead to C. And it’s based on the idea that the market is underpricing the near certainty of future earnings. It defies logic to think a cloud/SaaS with 95+ % retention rate will decelerate, etc.

But Stitch Fix requires a leap of faith - that if X-catalyst happens - growth will accelerate maybe even go nuts. There’s a greater degree of risk, of course, because the big catalyst may not materialize.

I personally feel the big catalyst will come. Maybe it will be a bricks-and-mortar component, new feature or ad campaign. Or maybe there’s just a thing where once X # of people adopt a service, growth goes parabolic. SFIX offers a better way to buy clothes than spending hours hunting through malls. Especially if you’re busy.

Anyway, I think this is an important distinction - SFIX is a little further from certain than the beloved Cloud/SaaS stocks here are.

The question I discussed recently with AJ is why invest in what might be when you an invest in what is? I guess one hopes that being in early on an event you’re confident will happen will offer a big reward. The way I see it SFIX is so well run that worst case scenario is modest underperformance, while best case is mainstream adoption. And the brand is so well managed it would be a fantastic jewel in many crowns - costco, wal-mart, target, amazon to name a few.

Best,

BD