Discussions about the fashion industry thread

[LA Guy]

Clients are waking up to the fact that digital is a tool, (a powerful one at that), amongst many. When digital starting coming in big into the agency and marketing world, a lot of money was waisted on clients buying and being sold shiny objects that weren’t right for their brands and businesses. Clients like Adidas recently admitted that they over invested in digital. Smart marketers like P&G are reinvesting monies into TV and Billboards which drive big ROI. These platforms and channels need to work in a connected way with each other to create a bigger experience. There is no silver bullet. There is no guaranteed best practice. There is no certainty. If someone tells you there is...they are full of ****.

This. Yeah, it's a useful tool, but it is maybe a bit too sexy, and it's easy to get seduced into big spends because the proliferation of data give people an ungrounded confidence in their ability to control the results.

And yes, company-by-company. I think that anyone who says that there is a general "best practice" is either stupid or full of ****, or very possibly, both. Conversely, there are lots of general "this is a terrible practice" advice that is good, lol.

 

[LA Guy]

Anyone who thinks they can apply "spend this here, spend that there" to every business seems like they're ignoring how every business is likely different. If you're just randomly buying Google ads, what can you expect?

Yeah, this is dumb practice if only because a $10MM spend and a $5K spend have nearly nothing in common. Also, as I stated in my last post, and a few before that, as has @London - it's really company by company, and a lot of it is still an art.

I think that humans just compulsively want "systems" to work. It takes the risk out of things, and you can be successful without an element of luck, which, let's be honest, would be great.

 

[LA Guy]

One of my favorite corporate conspiracies is the idea that there's a cure for cancer but companies are sitting on it because it's more profitable to treat cancer. *****, if there were a cure for cancer, the current CEO would take it, ride the sales while collecting the ridiculous bonuses, and retire the day the patent expired while flipping off their successor.

it's even better than the "Scientists made up Global Warming" conspiracy. Hell, we can't even conspire successfully on where to go for dinner. I've had many a tepid dinner in San Francisco because of this.

 

[Todd Shelton]

I would agree with you, at least on the substantive parts of your answer. I do think that statistics are better as checks on poor decision making or as a initial discussions starter, but I've rarely seen a good example of "data driven decision making".





Sorta. I am cautioning against the blind use of statistics and data analysis in general.

I think that as a general take home, my advice would be to spend more money on people and FTE, and less time and resources playing the numbers game.

Just to put my money where my both is, here is something that we have done, and then an example of something I've seen done by a pretty successful small company in menswear.

1) Styleforum buys google adwords. However, as we sell nothing, it's futile to compete with companies who sell stuff. They can afford much more for say "Saphir" or "Saphir shoe products". Plus, those search terms are generally used by people looking to purchase, or at best, read reviews prior to purchasing. Since we sell no Saphir products, we can't compete with those buyers. We also don't want to buy adwords on terms where Styleforum already scores high organically, and where the engagement metrics (time spent on page, pageviews, return visits) are already good. So, instead, we buy long tail searches that would be used by people looking for advice or discussion on the use of Saphir products, and where Styleforum scores high organically, but where the engagement metrics are low. In other words, google directs them to us, but whatever content they click on is not useful, in a bad format, or is simply the wrong page in a rather long thread. So, maybe "Saphir shoe renovation results". We then direct them to a specific, useful post, or an article with the user generated content put into a more usable form. We do look at the numbers again, but really, to see if the articles are worth investing in, or whether the forum post is just as, or more, likely to result in engagement. Sometimes it's a basically a toss up, in which case, we do the less expensive thing, and send peoiple to the post, other times, the articles are less engaging, so that's a dumb use of money, and sometimes, the articles are "better", at which point we have to decide how much "better" things must be to spend resources makiing articles. It's never clearcut, and a lot of human judgement is needed.

2) There is a company that invests quite a lot in forums, outreach to enthusiast sites, reddit conversations and subreddits, etc... This is where they engage, re-engage, and re-re-engage customers. The content, unlike a google adword, is pretty much permanent, and it keeps on reaping dividends. To support this, they have a fairly large group of sales persons, who are given X number of existing customers, and Y number of leads, with notes about their past purchases, their preference in color, fit, and category, and also, their prefered mode of communication, whether that be a phone call, a text message, Styleforum's private conversations, reddit's conversations, whatsapp, google hangouts, facebook messenger, etc... when say, new shirting becomes available, this is broadcast on the various forums and discussed, and at about the same time, the salespeople go to work, "Hey John, we just got some new chambray in, I think that you might like the PoW one, and it would go with your blue grey suit from a few months ago. You interested? Maybe you'd like to get a swatch and then decide? The swatches are $5 just becaue of shipping, but you get that back if you decide to buy just one shirt, so it's a good way to figure out if you really like it." It's high touch, not very numbers sexy, but it does make use of technology in a smart way. I think that they also do some advertising on Instagram, to get new customers, and when those customers have made a purchase, or maybe if they are somewhat hesitant to buy from a company sight unseen, they are directed to say, a Styeforum thread, or a @dieworkwear review, or a subreddit, or all of the above. Everything works well in conjunction. It's a smart, well conceived, and rather successful strategy. They are not ignoring data analysis, but it's nor really "data driven decison" either.

Solid reply, thanks. I agree with your thoughts.

 

[London]

The costs of digital customer acquisition are also going sky high on Google and Facebook, which have basically disintermediated a lot of clients from their data in the process. You see DTC's crashing and burning because of these costs. They are trying to 'create their own' communities because they are throwing VC cash down a sinkhole with he promise that growth will lead to profitability. That route works for very few.

 

[jet]

re: amazon I actually love the path they are heading down as it has tremendously helped small business

the amount of ads and business ideas of re-inventing everything we consume in an effort to sell **** is comical too I hope they continue to keep that coming, we are living in one big advertisement these days

 

[cb200]

we are living in one big advertisement these days

 

[cb200]

New (to me) DTC brand "Italic" promises the normal thing that DTC brands to be promising. What I notice is that in their online listing they are naming other brands that are sourced from the same factory. I've not seen this done before quite so boldly. It's a near perfect play for a value seeking consumer to frame the products in this way. Listing what the piece could sell for with a retail markup shows a savings that is only theoretical, there's no wholesale partners here so this price difference is made up. But it anchors the higher price as a value against the value. Using the other brands name implies that the pieces are equal to those of the name brands and doubles down on this tactic.

Final stage of the online DTC value play of disintermediation is cutting out the "brand" and buying direct from the factory.

 

[London]

The costs of digital customer acquisition are also going sky high on Google and Facebook, which have basically disintermediated a lot of clients from their data in the process. You see DTC's crashing and burning because of these costs. They are trying to 'create their own' communities because they are throwing VC cash down a sinkhole with he promise that growth will lead to profitability. That route works for very few.

Great piece on the challenges of the DTC model and how they are evolving:

outline.com is available for purchase - Sedo.com

outline.com

 

[cb200]

"The reality is that Thanksgiving weekend has become a time when retailers treat their worst customers like royalty."

Black Friday is bad for business

They say the customer is always right—but what if you're catering to the wrong ones?

qz.com

 

[smittycl]

re: amazon I actually love the path they are heading down as it has tremendously helped small business

the amount of ads and business ideas of re-inventing everything we consume in an effort to sell **** is comical too I hope they continue to keep that coming, we are living in one big advertisement these days

Fascinating read in the Atlantic on Amazon's warehouses and the injuries they cause. I worked for UPS as a sorter for a while in college and it was like working for the Gestapo.

Ruthless Quotas at Amazon Are Maiming Employees

Speedy robots. Constant surveillance. Ruthless quotas. A new investigation into why the injury rate for Amazon’s warehouse workers is sky-high.

www.theatlantic.com

 

[jet]

I know lots of guys building fulfillment centers at an alarming pace all over california. Hopefully the rapid growth comes back to bite them ********** like i predict.

Also they should have to start paying an ecological tax for all their bullshit.

 

[Nbarbar]

Why do you predict it biting them **********? Move to local fulfilment ie Local whole foods for amazon?

I know lots of guys building fulfillment centers at an alarming pace all over california. Hopefully the rapid growth comes back to bite them ********** like i predict.

Also they should have to start paying an ecological tax for all their bullshit.

 

[conceptionist]

Thought I'd chime in on the discussion on online/digital advertising and data analysis.

I have a graduate degree in statistics and have worked at a major digital advertising agency that employ PhD level data scientists/quants/statisticians. In general, large agencies and companies have in-house research teams that do genuinely good work, of which some gets published in scientific journals. With "large" companies, I mean the likes of eBay, Amazon, Google and other tech giants whose revenue are partially driven by ads, but also mid-size to large agencies.

Digital advertising is actually a very hot research area at the moment. One of the more famous empirical findings have been that most reported ROIs of digital advertising are grossly inflated. The reason is, as some other user here that agencies commonly bid on brand keywords so that previously organic traffic that was unpaid for, and hence not reported as a return on advertising investments, gets attributed to the advertising investments. Link to paper

To me it seems that much of the problems that relates to the hype of data analysis and the "science" in digital advertising can be explained by that digital advertising has not historically been a field that relies on empirical analysis or employ scientifically trained people. Industries like insurance, medicine and even finance have a much longer history of relying on scientific data analysis and employing PhD's, and hence, have much stronger infrastructure for doing "good" work that relies on data. As a result, in advertising there is a mismatch between the potential of the work and the actual work delivered due to that managers are often not sufficiently trained for the task, and in many cases do not even understand what can possibly be accomplished. However, I think it is reasonable to assume that the quality of the work will improve as the industry matures.

 

[London]

Thought I'd chime in on the discussion on online/digital advertising and data analysis.

I have a graduate degree in statistics and have worked at a major digital advertising agency that employ PhD level data scientists/quants/statisticians. In general, large agencies and companies have in-house research teams that do genuinely good work, of which some gets published in scientific journals. With "large" companies, I mean the likes of eBay, Amazon, Google and other tech giants whose revenue are partially driven by ads, but also mid-size to large agencies.

Digital advertising is actually a very hot research area at the moment. One of the more famous empirical findings have been that most reported ROIs of digital advertising are grossly inflated. The reason is, as some other user here that agencies commonly bid on brand keywords so that previously organic traffic that was unpaid for, and hence not reported as a return on advertising investments, gets attributed to the advertising investments. Link to paper

To me it seems that much of the problems that relates to the hype of data analysis and the "science" in digital advertising can be explained by that digital advertising has not historically been a field that relies on empirical analysis or employ scientifically trained people. Industries like insurance, medicine and even finance have a much longer history of relying on scientific data analysis and employing PhD's, and hence, have much stronger infrastructure for doing "good" work that relies on data. As a result, in advertising there is a mismatch between the potential of the work and the actual work delivered due to that managers are often not sufficiently trained for the task, and in many cases do not even understand what can possibly be accomplished. However, I think it is reasonable to assume that the quality of the work will improve as the industry matures.

We'll see. It's advertising and not science. You can't the human hand out of the equation and humans are biased. Biases are fed into algorithm design. Research is biased based on the design of the study and analysis of the info. I think the data/digital folks fed clients a lot of snake oil, because they knew that they weren't very educated in the field and they equated digital with innovation. Advertising was heavily weighted towards the art side. The digital guys came through promising attribution and measurement clients went bonkers to measure everything under the sun, even though a lot of the metrics, where vanity ones, thus swinging the pendulum.

A lot of clients wasted a ton of money on shiny objects that had no material impact on their business with the promise of 'being digital." I've worked at a ton of agencies and some of the best digital ones. I'v seen it up close and personal.

Clients are waking up to the fact that performance marketing has fractured their brands into a million pieces and led to over-investment in a lot of people that they're already reaching. They forgot that the job of the marketer is to differentiate, build brands, and bring more folks into your tent.

A lot of clients are waking up to that fact. They are investing in so called traditional channels like TV, billboards, etc that can't be ad blocked or laden with fraudulent ginned up metrics.

You'll see the pendulum swing back and become more balanced in the media mix.