When was the last time anyone questioned the financial outcomes of retention marketing activities?

Going by the logic:

Every action you take must have a future expectation of results...

Nothing should be left to best practices or "experience".

This is the baseline of all good marketing.

And no one should ever settle for less.

Yet this is pretty typical for many agencies...

Trust-me-I'm-an-expert "strategy".

Nurturing customers, meeting the customer where they're at in the journey, personalisation.

Makes sense in theory, but what’s the incremental lift?

Is it the most efficient use of your limited resources?

At the 7-figure mark, trying to break 8-figures, this is an ever more crucial time.

That's what CTC's Global Accelerator Program solves for.

1. Better alignment with traditionally siloed growth <> retention teams.
2. Better agility to support a very dynamic marketing calendar.
3. Better analysis of activities that impact your P&L.

Together with Growth Strategists, Retention Strategists make this work.

At CTC, we always sign our names on the bottom line.

Because we know that’s what it really means to win.

If this speaks to you, let's chat and see if we'll be a good fit for each other.

Josh Tay

Some good stuff from this month.

Because everyone still asks to see designs. So here we go :)

Director of Retention, Common Thread Collective | Building the world’s strongest retention program for 7 & 8 figure DTC brands

Just because you sell skincare, doesn't mean you need to only talk about skincare.

In this example, we tested running traffic to 2 quizzes.

Quiz 1: Which smoothie matches your personality
Quiz 2: Which beauty routine is right for you

We didn't expect the smoothie version to outperform by so much.

Most people would've stopped just after creating a quiz.

Or maybe compared quiz AOV/LTV vs non-quiz takers.

But we thought this would be a worth-while test.

Lesson: Think about your customers' lifestyle. 

And alternatives (not complementry) for your product. 

Weave them together.

Now this is a fascinating insight that our retention team passed to the growth team.

Client industry: Apparel

And we found that customers who bought designs with dogs had much higher LTV than those who didn't.

What brand will be doing:

- Add new filter for dogs for easier nav
- Test new designs with dogs

What we'll be doing:

- Dedicated dog ads
- Find more of such niches

How much time is your team investing into drawing insights?
Is your team reporting on metrics or taking action?
Are you asking the right questions?

Giving discounts in your flows to increase AOV?

What criteria do you have to determine who receives a discount?

Everyone?
Non-customers only?
Discount with threshold? But which amount?

To answer that, you'll need to look at your AOV histogram.

1. If you based it off AVERAGE order value, you might be pretty off

- Averages lie
- If we had set it at the mean, there'd be a very big chunk of people who wouldn't have moved
- Purely because their cart value might've been too far (see how far mode is from mean)
- Mean is getting pulled up by select customers who spend enornmous amounts
- Question is, where are most of your customers?

2. Look at MODAL order value for help

- Your objective is to move people one bucket up
- Moving the bulk of customers up will translate to higher AOV by default
- It's the same for how you'd create free shipping thresholds or free gift with thresholds in your gamified cart drawers.

There's math to most of these things.

Only do things when it makes sense for the business.

You might win on revenue, but at what cost?

(Long post ahead!)

Today I want to breakdown a real post-mortem of a recent play we did for women’s headwear brand.

And exactly how we still hit the forecast while mitigating downside risk. 

Context:
Client wants to repeat last year’s promo mechanics and marketing calendar.

Expectation:
Client looking for 1.5x rev growth with 20% higher CM.

So unfortunately, that means getting more out of email and sms channels

Problem:
Based on projections following last year’s cadence, we weren’t likely to hit target revenue desired by client.

Constraint:
At the end of the day, it’s the client’s business, and we’re here to deliver results. So it’s our job to best solve for it.

Is this sustainable? 

Nope. But it’s the best stop-gap measure before we have a conversation on expectations, while still delivering on performance to prove we can.

👀 Observation of historical:

Every major promo (BFCM, Mother's Day, Labor Day, Independence Day) did generate more revenue YoY. 

Revenue up 87%, but they had to send to 133% more people to get there. 

At some point, we’ve just been adding subscribers who don't convert into the campaign mix…

Dragging down the economics of the entire channel.


🧠 Insight 1:
RPR declined every single time—ranging from -7% to -42%. 

What was happening is that they were running the same plays to bigger lists and getting diminishing returns per subscriber.

It was reusing the same promo mechanics promo on promo, because “it worked”.

And expanding audiences toe extract more revenue.

What was “working” was going to crash real soon.


🧠 Insight 2:

The correlation between campaign frequency and RPR was actually positive (+0.25).

More campaigns in a month didn’t tank performance. 

What does correlate with declining performance though, was list size growth:

Early period: 56K avg recipients/campaign → $0.0326 RPR
Late period: 126K avg recipients/campaign → $0.0261 RPR

As the send sizes expanded 2-3x, engagement metrics collapsed.


🔬Analysis:

Most would think we have a list fatigue problem.

Initial thought is to reduce sends.

But we actually don’t have a demand elasticity problem.

We are just dealing with the law of shitty cohorts.


👎 The wrong things to do:

1. Expand audiences to get more impressions.
(This will trigger law of shitty cohorts)

2. Reduce send volume
(This wouldn’t allow us to hit rev targets)


👍 What to do:

1. Increase send volume to tighter engagement segments since list has been proven to be able to handle.

2. Product promos only to go out to product-interest segments.

This is why we never skip forecasts.

The forecast allows us to gap-find and it’s based on the exercise of rigorous forecasting that we uncover these insights that we’d otherwise have missed out on, which might have resulted in incorrect execution.

When creating any post-purchase upsells/cross-sells, always rely on your product journey map.

The common way:

Person adds sample > They buy regular next

Data might show otherwise.

In this case, person goes straight into bundle.

Even for singles, they seem to not replenish the same product.

Instead, they gravitate towards bundles, and stay on bundle orders.

This could drastically change the way you push for repurchases.

Interestingly, this mystery item was the fourth most purchased item.

This item's added to the cart drawer as a default upsell on every product.

Avg item price: $35
Mystery item: $15

Objective: How can we liquidate slow/dead inventory faster?

Go-forward questions:

1. Should we create a mystery bundle option too?
2. Can we create "mystery gift" as a front-end offer?

These will all be tests to be conducted.

Ask the right questions...

... Come up with the right hypotheses

... Test the right things

... Take the right actions


Client niche: Accessories

Klaviyo audits alone are insufficient to determine what to do for your retention program.

How should you prioritise activities?

Of the flows recommended, what will the content/angle be?

It's one of the reasons why we look through everything.

Based on this finding, we're:

- Doubling down on resources for refer-and-earn program
- Creating multi-bundles and quantity break discounts
- Include more creative for influencers in pre-purchase flows

(we found that most people who said instagram, youtube, tiktok, also meant influencers and creators)

To do a sanity check, we looked at codes redeemed and saw that most first time customer codes redeemed were from affiliates.

Sometimes things like these we're overthinking.

But we're actually seeing higher lead:conversion efficiency across the board in the last 90 days vs previous period.

So we know that's working.

---
Client niche: Fitness equipment

4 mistakes I made when first forecasting email revenue and how you can avoid it too.

I wish we could figure this out faster too.

But the retention scape in this aspect is still in its infancy.

We're doing our best to fail faster.

To come to a clearer truth.

Here's 4 things I got wrong and how we're solving for it now:

1. Not taking monthly volume into account (rev degradation)

❌  As we increase campaign volume, unsubs might increase and RPR might decrease.

What should've been done was to create a weighted average of RPR based on volume of sends per week over time, split by promo vs non-promo

---

2. Not taking post-promo revenue dips into account

❌ From the last promo sent, you need to determine the "cooldown period" before demand picks back up again.
---

3. Not taking last-year-same-period's email volume into account

❌ For the sake of simplicity, I naively assumed last year's RPR to be same as this year's. 

I just applied a seasonality factor on top with a YoY growth factor.

But the diminishing returns from higher volume played a bigger role than I anticipated.
---

4. Not discounting active subscriptions in email campaigns

❌ If you send a general educational campaign to your engaged list with active subscribers in there...

You'd have been inaccurately attributing revenue.
---

Now we're factoring all that and building it into our forecasting algorithm.

And fully automating this to eliminate human error.

Humans are terrible at extrapolating data.

Worse if we need to incorporate several other aspects.

Tech enablement is the future of marketing.

Built upon solid reasoning, 
asking the right questions, 
with an endless desire to improve, 
and relentlessness to follow-through on the outcomes we promise on

Your cart drawers, checkout upsell, post-purchase upsell funnel will bump your LTV up more than a bounce-back post-purchase email will.

On these pages, you have 100% open rates.

With the highest purchase intent.

Take advantage of it.

Different products in cart might need different upsells based on:

- LTV growth by product
- Total LTV by product (including AOV)
- Balance with product demand (volume)
- Product journey
- Frequently bought together

Then once you have your various funnels/upsells in place, there should be monthly measurement to identify what needs improvement.

How equipped are you and/or your retention team right now to execute this?

Reviews immensely help reduce friction to the purchase.

People look at the number of reviews to know if this brand/product is trustworthy.

It makes sense to build your review war chest.

In post-purchase surveys, we've asked specifically about reviews.

Question 1: "What made you choose to shop with us?"

Options: "Reviews", "Referrals from friends", "Video tutorials", "Others"

Of those who selected Reviews, we had a logic to ask the next question...

Question 2: "How did the reviews help you make a decision today?"

Options: "Other customers' stories", "Photos/Videos of products in use", "Number of reviews vs other brands"

And we saw high photo/video review responses.

This isn't a theory propogated by review apps.

We personally tested this.

That's how we know what to do.
(Always be skeptical and own your own data. Everyone on the internet lies. Maybe I'm lying too)

But if you're too lazy to test, here's what you can do to set up.

1. Go to your review app
2. Create coupon tiers
3. Customise review request emails to let customers know about the added incentive
4. Set an expiry on the discount
5. Send reminders to customers who left respective reviews, to claim their discount

Optional:
A/B test with Intelligems or a similar tool, with and without reviews on your PDP.

Or with the number of reviews stars right above your product name.

And see how that affects conversion rate.

Who said retention was only emails and sms?

Is your retention person supporting you with these?

Always nice to know we're pacing on target, within <10% delta.

We still have big moments we're expecting towards end of month.

Tracking performance this way takes away guesswork.

The worst thing an agency can do is to come up with excuses to justify underperformance.

Over and over again.

When all that was needed was accountability.

We're now more than 90 days past BFCM.

You know what that means?

Time to reactivate your black friday shoppers who are unengaged on digital.

Looking at your repurchase windows, you'll see that majority of LTV lift for new customers happen within first 90 days. 

(typically)

If you're a brand with replenishable products...

Or you sell things with high novelty/"new drop" factors...

This is definitely something you'll want to get in on.

Relook at past year's data and you'll also find this:

Your black friday buyers are a lot more sensitive than other promo period shoppers/evergreen.

(I've had a post with some of this data before if you wanna check that out.)

So if we're nearing the end of the highest probablity curve of repurchase and they're still not coming back...

Much less engaging...

You'll want to do something about that.

1. Offer
Again, we know holiday shoppers typically only buy with discounts

2. New channel
If they're not engaging on email, sms, try new mode of communication

And this is the perfect time to start.

+ you only have a very short window to execute.

You'll want to run your winbacks for this group of customers before the next major sale.

Which for most US brands, is going to be Memorial Day and 4th July Sales.

If you start right now, you'll need a couple of days to get post cards delivered to doorsteps.

Then, people don't take action immediately.

Postcards might sit on their desks or coffee tables for a while.

That's why we wait out with a 60-day maturity period.

🚨 IMPORTANT:
You don't want your winback offer clashing with your site offer.

Because now that just muddies data and the offers cannibalise each other.

Timeline:
Now > 14 Mar = Plan your campaign, offers, creative, segments
15 Mar > 20-ish Mar = Delivery
20 Mar > 20 May = 60-Day Maturity Period

Just in time for memorial day promo launch date.

And you can do your usual holdouts.

But this is a crucial time to strike.

If you want to run a proper winback campaign, do it once, do it right.

Have an Amazon store that gets more traction than your .com?

Consider an exit intent popup redirecting traffic there.

Targeting:
- triggered on PDPs only
- $0 cart value
- stayed more than 15sec
- or scrolled more than 50%

Create specific popups for each key product, with relevant redirects.

Instead of wasting that visit, send them to somewhere you know people are actively converting on.

We're doing this with a home improvement brand that sells DIY light kits right now.

If people want to shop on amazon, we can't force them not to.

Return rates on amazon for this brand aren't crazy high.

So this will be our interim solution before the growth team solves for the conversion issue.

There's no one way to use popups.

Everything is just a tool.

What matters is how you use it to benefit your business.

“Bundles increase AOV”

Logically, that should translate to higher LTV.

So many people default to bundles, which might not always work.

In this case, I’m specifically talking about pre-set bundles, which is the most common way of bundling. 

Sometimes customers just want to pick out exactly what they want.

Forcing a bundle may actually end up reducing conversions.

Ideas like:
Hero product with slow movers
Hero with high LTV product to push profit
Hero with accessories that go well together

These are valid pairings.

But again, we’re assuming people want the next product we’re recommending.

What you want to sell might not be what customers want to buy.

There are two ways to go about this.

Product journey + frequently bought together analyses

Which I’ve talked about before in previous posts.

Let people choose on their own

Aka Build Your Own Bundle.

This works especially well (and much easier to implement) for brands with similar AOV products in the catalogue.

And you can gamify this with the typical tiered discounts with threshold.

I’ve seen bundle take up rate go up to 20% with dedicated BYOB pages in the nav bar.

Example 1 
We have a client selling canned tea. They have 12 flavours. Instead of forcing 6 packs of single varieties then force customers to buy multiple carton to try different flavours, you select your own quantity of flavours for each 6-pack.

Structured like a quiz

Step 1: Select how many cartons you want
Step 2: Carton 1 - Which flavours do you want
Step 3: Repeat for however many cartons selected
Step 4: Would you like to add another carton for X% off? (Progress bar of discount at top of page to)

If AOV varies widely, adopt a similar strategy as how you’d buy a computer.

Base product
+ one add on from category 1 at $X
+ one add on from category 2 at $Y
+ one add on from category 3 at $Z

Each of these are optional.

Example 2
Another client selling sofas. That’s base product. Category 1 are removable spill proof covers. Category 2 are pillow sets. Category 3 is a 2-year warranty.

In our frequently bought together widget below, we still have opportunity to cross-sell sofa extenders and poufs.

Now you give customers lots more flexibility while maintaining some level of control of options.

No fluffy customer journey theory.

Just real stuff that you can implement right now.

And all the data points you need.

Feel free to take this and run with it.

Client niche: F&B, subscription

How to overcome weak subscription program take-up rate

"Multi-month" cohorts almost always will have higher LTV than monthly subscription cohorts.

More specifically, I’m talking about cumulative LTV.

Which is at the end the most important thing to the business. 

That $ value.

So this is definitely something every subscription brand should be testing.

In this example, brand sells soaps.

- One-time orders
- 15% monthly subscribe and save
- 19% for double pack bundle
- 23% for triple pack bundle

Your CAC might be higher on these multi-packs.

Largely also due to much higher AOV.

But not only is cumulative LTV higher, margin LTV per customer beats monthly subs hands down.

Essentially you’ve collected ltv upfront. 

Which is great. 

You don’t have to worry about churn as much. 
CAC payback’s shorter. 
So is your cash conversion cycle.

Definitely a test worth trying.

LTV comes in many forms.

Do what's right for your financial outcome.

Are you levergaging Web Push notifications?

- Welcome
- Browse abandons
- Cart abandons

If you find your lead:conversion rate on these email/sms flows are weak, it might make sense to have another channel to support.

At the end of the day, it's about impressions.

So where do your customers hang out, that you also be on?

Show up there and drive them back to site.

Then track lead:conversion efficiency

The way you structure a tiered loyalty program must take psychological triggers in mind + factor in unit economics.

Weak participation rate = Poor incremental revenue

= unnecessary complexity 

= useless extra cost to run the program

Three post-purchase micro surveys conducted across three different brands.

All had one same option:

"When there's a sale on"

In all instances, that came up as the highest volume response.

This is probably the reality for most ecommerce brands.

People are most likely to shop when there's a sale or a discount.

Knowing this, it's ever more important to bake the discounts already into the price of your product.

So that you can afford to "discount" when necessary.

Without having a really strong, notable brand, is it really possible to never go on sale, and still maximise your growth potential both on new and returning rev?

What do you think?

SuperAGI

Growth & Revenue

Contlo

Forward

Founding Member - Sales

Vajro

Sales Lead

Sr. Account Executive 

Account Executive

Pickyourtrail

- Closely worked across cross-functions like operations and customer support.
- Was part of the Paid campaign, to help bolster revenue from Google paid leads.
- Worked across multiple destinations in SEA ,and helped scale smaller destinations like Bhutan and Sri Lanka.

taplio

The Battle

Josh Tay

Gary Thomas

Followers

Posts

Engagements

Likes

Performance ScoreJosh Tay's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Josh Tay

Gary Thomas

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Josh Tay's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreJosh Tay's LinkedIn Performance