Home > Paid Traffic Sources > eCommerce

---

3 Guys Conquer the Beauty World (49)

---

Very interesting. Thanks for sharing. Any reasons why you guys got into skin care instead of something like diet pills or mobile or something. Just curious.

---

Thanks @melachii

Niche was chosen based on the engagement and research. E-commerce instead of mobile because we have a lot of experience from e-commerce.

If we make this project profitable in the next few months we will probably go into subscription model and/or iOS app.

---

#Ibelieve. Nice work so far App! Looks like you are going to crush it if you keep it up. This should be a good thread.

---

Thanks jasonaao.
On 3/25/2016 we'll start pushing traffic from Facebook and Adwords to our product sets and in the meantime we need to make some sales-letters, product emails and Shopify modifications.

---

Very nice work, I did e-commerce business(selling replica clothes, shoes, CDs) few years ago, it was hard to make sales at that time, we only rely on SEO traffic due to lack of marketing knowledge.

hope you guys make profit sooon....

---

Thank you for your comment cnpabi.
I think we're pretty lucky because we have a lot of PPC and email marketing knowledge. I'm not sure if we can make w/o paid traffic...

---

Very insightful and motivational. I'll be following this thread and will not be surprised to hear that this project has turned out successful for yourself and your team. I'm sure all these steps you have taken will pay off when you launch your products on the 25th March.

All the best to yourself and your team

---

Thanks lachz17 thanks for your words of appreciation.

Quick update

Video Content
We gave all our products to a local girl who is using products for professional use. In return she will create 30 short-videos (15sec). One video is already published and engagement is surprisingly high. We will put a lot of our effort in "video-marketing" because we are seeing huge potential for promoting our products on this way. I will keep you posted about the progress.

Reports
At the very beginning we want to have an easy overview about profitability for each product in combination with traffic source. Temporarily solution for this will be customized Google Spreadsheet for export from Facebook and Adwords (labes/tags). Please let me know if anyone is interested for reporting spreadsheet.

First order
We're waiting for the first "real" pro forma invoice. First order for $3000 will be sent to fulfillment center later this week, if everything goes as planned. Fingers crossed.

We're waiting for more questions from you guys.

---

Heya guys

Would be keen to jump on a mastermind with you all. I've been doing womenswear ecommerce for the last 6 months and have done around 200k GBP rev, 95% driven from FB. I've just gotten back into office after meeting some suppliers, but will post some thoughts later today.

cheers

---

Hey black hand!

Awesome, congrats for your e-commerce project.
I'll send you my Skype and let's get in touch!

---

Update 3/18/2016

Everything going according to plan and our timeline.

This week we're polishing webpage (credibility, UX, design), editing products (SKUs, GTIN, MPN, descriptions, images) and working on banners for pilot campaign on Facebook.

Upwork
We outsource some UX/design tweaks and find one super-responsive girl on Upwork. We gave her two quite simple test jobs and she delivered us great work. If someone need a good contact for Shopify developer ($22/hour), please let me know.

Calculations
Now that we have all costs (fullfilment, COGS, transport, ...) we made some forecasts.

Clicks = 1250
CPC = 0.35
CR = 0.80%
Advertising Cost = $437.5
Orders = 10
Revenue = $1,290
Net profit = $309

I'm quite sure that we can get CR 0.80%, actually I think it will be much higher and with really targeted ads CPC in US should not exceed $0.35. If this forecast hold water then we make $309 Net profit at $437 spend.

We're all totally excited about the launch but we need to go step-by-step and stick to the plan. This weekend the whole team will brainstorm together about how to develop subscription model because we have an idea that we could make retention with subscription model. We'll give you more details on this after the weekend.

Have a great weekend!

---

Update 3/22/2016

We're are two days before launch. I hope that our next update will be about how many order did we get.

First order
Yesterday have made payment for our initial order and put deposit on our fullfilment account.

Google Merchant Center
We need to setup properly our product feed inside GMC if we want to run Google Dynamic Remarketing. There was an issue because our fullfilment center (Shipwire) have Code-128 instead of GTIN so we need to convert Code-128 in the proper format for GMC.

Sales Funnel
Most of our time and energy before launch will be put into testing the sales funnel. We need to modify some workflow inside MailChimp and test the triggers/links/tracking etc.

Custom packaging
First 50 customers will get our products in the very basic packaging w/o our design and logo. We're aware that this is super-important for user experience. We're already working on the packaging and looking for custom made packaging supplies. Does anyone have any recommendations maybe?

CR optimization
We want to make conversion-rate optimization as part our daily routine because we're aware that already a slight CR improvement can have very positive impact in ROAS. At the moment we're in the research mode what is the most flexible and powerful solutions for this. I would appreciate any recommendation.

That's all for now!

PS: Is there any specific reason for such a low response on this Follow Along? I guess we need to move some of our advertising budget on native-advertising.

---

Launch day! Looking forward to the update

---

Update 3/27/2016

Launch
Launch was as planned on 3/24/2016 and everything went well. On day1 we spent $350 (Facebook, WC (A2C), 10 interests, carousel, female only, 18+) on sending traffic to product page for one of our products (regular price: $129) and bounce rate was 93% which is totally unacceptable. Reason for this is probably the price. This week we'll do some split-tests to get more information on this and analytize the reports we got from usability testing tool - peek.usertesting.com.

We shutdown FB WC campaign and enabled FB Leadgen, here are the stats:

Facebook LeadGen Camp (2 days)
Spend: $108.32
Cost per Lead: $0.45
Link clicks: 462
CPC: $0.23
Sale: 1
Revenue: $103.20

Leadgen campaign is working great right from beginning and we got out first sale! We decide to first establish first our email sales funnel with FB LG ads because we're getting emails, likes, clicks for really cheap.

This weekend we're planning to figure it out the reason for incredible high bounce rate on WC ads. Bounce rate from sales funnel emails is 30% which is something was we're looking for.

More later...

---

Awesome, follow along. ha first sale. congrats !

---

Thanks! We hope you'll find some useful information in our follow along.

I forgot to mention that we find one awesome service for connection between Facebook LeadGen ads and all popular ESP. The service is called: Leadstunnel. We're not connected with them in any way. Just a happy customer.

---

Congrats on the first sale! Subbed to your thread, this should be a very interesting follow along.

You can also check tools like Hotjar, VWO, FullStory, UsabilityTools and others for optimizing your site and conducting some split tests.

From personal tests I found that if you add even some basic imagery to category/product pages, even supposing expensive products, with a bunch of benefits and bullet points basically restating/reaffirming your ad copy can, in most cases, decrease bounce rates. Super simple stuff like that usually helps, even if you don't already have the data and tools to back your claims up.

Looking forward to more updates!

---

Great thread. Looking forward to more failed attempts before finally cracking that 1k/day mark (profit).

---

in for this

---

I am not a Zeno of FB, but I'm currently running profitable campaigns for a few ecommerce sites and although I haven't gotten to properly scaling the campaigns, here are some of my experiences so far, maybe they'd help:

- DPA ads, at least from my tests based on the products/targeting/angles, performed the absolute worst. No matter how many ad variations I test they just perform poorly - maybe it's because of the niches.

- CVR ads are performing the best for me right now, by a long shot.

- Whenever I test stuff I make sure to test every single possible ad type - CVR, Clicks, Leads, Engagements(these did well too), Image vs Video(video ads + conversions are killing it right now, for example) and so on. Also audience retargeting(a whole number of different custom audience types) tested with all the ad types. As you mentioned above DPA campaigns didn't produce any results, I had the same thing happen to me with a number of different ad types, whereas others were breakeven/profitable.

- ^ All of the above tested for every single angle I come up with, based on extensive research of the targeted demographic.

- Test manual vs automatic bidding vs scheduled lifetime budgets too.

- CVR ads were very, very tricky to get to work before I actually understood what was going on under the hood. For a short summary: you need to let the ads gather data for at least a few days, or ideally a week before you try to optimize or change anything(even if you have a million split-tests going on for different angles) so that FB can gather the necessary data in order to start sending you well converting traffic. E.g. if you plan on spending good cash, spread it out across as many days as you can instead of spending it all in just a day or two. It's much slower than the others and also takes a lot more cash out of your pocket, but it's worth it.

- For CVR ads, test every single different conversion goal you can imagine - from all the standard events to all the custom ones on your site, all of these going for the same time period I mentioned above. The difference can be quite huge. I usually just get some pageview goals up to let facebook sort out all of the data but have all the other conversion goals setup beforehand, this makes optimising later on much easier imo.

- When talking about CVR ads I test every single "variable" in separate campaigns - i.e. angles, images, ad types, bidding modes, all of these go into their own campaign, each of them having just 1 ad and 1 ad group. The data is much more accurate than having a lot of ad groups, because FB seems to prioritise the ad groups which it thinks are doing/going to be doing the best and this gave me a lot of skewed data, which was literally wasted money for uneven split-tests. This only applies to CVR ads though.

- I'm not sure how many products you have on your site, but I found that targeting ads at either 1 specific product or product category performs much better than trying to craft angles for the entire site's products and your general/broad value proposition. Relevance, clarity and tighter angles do better, as expected, I guess.

- Having VWO or something similar implemented is, I think, extremely important. Optimising your site as much as possible is vital. The typical ecommerce layouts/templates/themes and designs often times perform very poorly compared to what you can make them do if you put in the time to optimise the actual site. After I carefully studied the analytics of one of the sites I'm running ads for I saw it had a huge dropoff in traffic and super-high bounce rates on one of the pages, which after some optimisation now pretty much turned the entire thing on the profitable side.

In other words, I did the full-scale typical MVT testing of every single thing you can imagine and then you switch over to (insert fancy words here>)factorial design once you know what actually works when you get the needed data.

Hope this helps in some way, but again, I am by no means a FB expert and these are just personal observations and what worked for me based on low overall XX,XXX ad spend, so take it with a very healthy grain of salt. I'm just barely getting interested in FB.

---

Wow @shishev, thank you for this great post.

This is exactly what we're working right now. Try to find a combination of interests, bidding and ad type and try to scale up fast.
One of the mistakes we did was changing the campaign settings too early and because of that Facebook did have enough time for optimize the traffic. From today on we will leave every combination for at least 72 hours before taking any actions.

@shishev - I have also some questions

- Can you maybe tell more about video ads? We have some videos and bidding on CPV bring us cost of video view $0.003 which is ridiculously low I think. Now we're running retargeting all users who completed the whole video.

I cannot agree with you more on this. CVR optimization is one of the most important part. Because we're spending only $150 daily, we run only one AB test at a time. Currently we're testing product layout: a) product video in description b) w/o product video. As a goal we set click on add to cart.

---

For the video ads, I pretty much setup a split test for CVR type of ads (also add to cart goal at the time) - my best performing images vs a well crafted and interesting marketing video showcasing what this product can do in real life. All of this with the same ad copy/angle, just swapping the images for a video and so far it seems to be working much better. Though I got this video from the original manufacturer of the product and I believe they have spent a lot of cash making it.

Previously I tested another video, with sort of the same note to it, just not as interesting nor having that "wow" effect to it and it too did well, however it died out after a point but my main ad images (still running the split test) were still performing well. Really not sure what caused that particular video campaign to die out. Talking about conversion ads here.

The current video that's doing well has been running for a few weeks and has no signs of deteriorating anytime soon.

I also have that initial video (the one that died out with CVR ads) running on a tiny Video views campaign and video view costs are almost the same as yours - it used to be $0.002, and it's $0.001 per view now (I guess it settled down? ha). This one is purely for engagements/branding but although it has spent just about $300+ it is actually profitable for the lifetime of the campaign. Not only that but we also got offers from people wanting to resell the product and get it into their actual physical stores.

My personal guess is the cause for these videos' good performance is their overall quality - they address a lot of the pain points and show solutions in an interesting way. I definitely won't be able to produce something like that from scratch.

I also have a retargeting campaign that's doing great and getting good sales - something along the lines of targeting "people who have visited your site but haven't been to the final checkout page(or cart)" type of thing. This combined with a discount coupon seems to be working well so far, though it's quite a tiny test.

---

Updates?

---

yeah, in the green. well done. lots of great info here, thx for sharing the details !

---

this thread is epic.

just found this today while in the midst of starting the same type of project.

subscribed.

btw. what exactly does leadstunnel do? and did you find a program for FB and Shopify ROI and cvr tracking?

---

Loving the progress here, looks like you guys are on to something big here, keep up the great work!

---

Thanks man!

Leadstunnel allows you to automatically synchronize your leads from Facebook Lead Ads to almost any e-mail software provider - Mailchimp, Aweber, Get Response, Active Campaign etc.

Currently we're calculating ROI from FB and Shopify reports in Google Spreadsheets on a daily/weekly/monthly basis but we're looking for something more simple and flexible. At this stage I think it's better that we have a lot of manual work with those reports because we have a bigger picture about what products are performing and where to put money and focus.

We're looking for a way how to include COGS in Shopify report. Today I got recommendation for Profiteer extension. I guess we will try it and see how it works.

@kiwanuka94 CVR tracking is calculated in different places -> Shopify / Google Analytics / VWO. It depends on what you are looking for. For overall progress we're look CVR in Shopify, for optimizing ads campaign Google analytics and for CVR optimization VWO. Hope this helps.

---

boom. this thread is coming at the best time I have to say.

definitely interested in Cyfe.com.

What would you say have been the biggest factors in your uptick in success the past week or two?

---

We've used www.jukeboxprint.com for custom packaging probably not the best prices in the business but the product came out looking really nice.

---

Great! I'm glad you find this follow-along useful.
Cyfe is really awesome service. You can simple create graphs/tables/... directly from Facebook/Google/Shopfiy/Wordpress/ESP/Zend Desk/and much more. One great feature that is important for me is Google Spreadsheet integration because you can create custom reports easily and show then on the dashboard. There is available Google Analytics Spreadsheet Add-on which makes easier to manipulate GA data in Google Spreadsheets.

I would say that there was two factors that have the biggest impact on the ROAS in the past month:

1) Using VWO Site Testing Software
From the first day we're running different split-test and trying to optimize the UX and CR. We made huge improvement here. In one of my previous post you can see what did we test.

2) Only Desktop traffic from Facebook
Temporarily we're buying only Desktop traffic from Facebook. Every morning I take 15 minutes and I'm watching video recorded sessions from our webshop visitors. Based on this I found out the our UX issue on mobile and we're working on it. I hope that later this month we'll start buying also mobile traffic.

Thanks! Packaging on their webpage look great. I will contact them.

---

I can recommend a chinese company for packaging. Got some custom boxes done for I think around $2 for 1000 units.

---

Yes, give me more details please.

---

I'd love to hear some more details on the app you are developing.

Also, I would appreciate if you could give me a quick rundown on how to get people onto a custom audience after spending 5 minutes on the site.

---

Send me over your email and I'll give you beta access to the app next week.

Google Tag Manager -- Fire pixel after X minutes

1) New trigger in GTM
1.1) Name = Power Users - 5min; Event = Timer; Event Name = gtm.timer; Interval = 300000
1.2) Enable when = Page URL matches RegEx .*
1.3) Fire On = All Timers

2) New tag in GTM
2.1) Name = Facebook - Power Users; Custom HTML Tag
2.2) Configure tag =
<script>
fbq('trackCustom', 'PowerUser', {tos: '300'});
</script>

2.3) Fire On = Select the tag you created in 1 step (Power Users - 5 min)

PS: You can follow those steps if your Facebook Pixel is implemented over Google Tag Manager.

---

I'd definitely be interested to hear more about your Shopify App experiments - it's a market I've considered going into.

Great Follow-Along so far!

---

Thanks caurmen!

I'll open a new thread only for Shopify App and share our marketing plans and others useful data we got on our way. If everything goes by the plan we're going live at the end of next week.

---

Joining in late, but man.. what a follow-along so far. I love they way you guys pay attention to detail and how you put emphasis on tracking KPI's.
1) Any reason to why you chose GA over performance marketing trackers? (eg. Funnelflux, works with email funnels as well)
2) Would be cool if you could share your Google Spreadsheet templates for calculating various KPI's.
3) Didn't know you can record screens with VWO them, checking them out How much success (decreasing bounce%, increasing CR% etc) from all different factors would you account to optimizing UX and eliminating friction.
4) Not familiar with Shopify, we've built our funnels with Clickfunnels. You build your sales funnels with a Shopify app?

All the best!
P.S. Thanks for Cyfe.com, trying it out!

---

@Mr. App - Fantastic follow along. I just started my first project in the e-commerce space and you've probably saved me some strife. Gotta couple of questions for you:

* You mentioned looking at bundles to push up AOV. Are you currently doing any cross-selling / upselling, or is all rev from single item purchases?
* Looks like you turned the profit corner at a sale price closer to $33 rather than the original $129. Change in offer?

Reasonj for asking - The product I'm testing ATM has a sale price of $19.99 but I may raise it to $29.99 shortly. No upsells or bundling currently, so I'm unsure if I can drive CPC low enough on a single unit to turn green.

---

Follow along of the year! Keep it up!

---

Great progress mate !!

I sent you a PM too but I understand you got many, so I'll ask here may be someone else can jump in.

Does mailchimp force the confirmation (double opt-in) for leads collected via leadstunnel on FB lead ads ?

They say so on leadstunnel .. and that's kind of a big let down if it's the case.


Mehdi

---

Did you use aliexpress as your supplier ?

---

Hey Mehdi!

Sorry about the delay in my response.
We ran one month ago FB LeadGen > LeadTunnel > MailChimp and I did not see double opt-in issues.

I will re-check again today and let you know.

---

No worries mate

I must have missed the part where you switched from FB Leadgen Ads, I thought you were still only using those.



Mehdi

---

Amazing thread, very motivational. Any update?

---

Fantastic follow along! Really amazing! Move on! If you have any problems with working with Chinese companies, you could message me. Maybe I could give you some ideas

---

Any updates?

---

guessing they are doing quite well...

---

Interesting, follow along. Good luck to you.

---

Hey

Great follow along for your Shopify store. Hope all 3 of you guys are still doing great.

Personally, I'm no expert in Shopify but good friend of mine is in the same niche aka beauty products. He is killing it with stable minimum 5K a day revenue.

Most of his traffic comes from IG. This is what he does:

- He finds models on IG with min 50K+ followers
- Sends direct message with offer to send them his product
- They post the picture on their page with his product and tag his IG
- As a result 90% of his traffic comes from IG

---

Home > Paid Traffic Sources > eCommerce