Kristen Gilles

Post-Cruise Nudge:
Behavioral Timing & Emotional Design

View Email Here
This email was designed around a simple behavioral insight: memories fade fast, but motivation to record them follows a predictable emotional arc. Rather than sending a single generic "share your trip" prompt, this was designed as a two-part touchpoint strategy tied to where a guest was in their emotional journey, a mid-cruise email, sent while they still had time and energy to reflect in the moment, and this post-cruise nudge, timed a few days after return once the logistics of coming home had settled but nostalgia was still fresh. The subject line ("Before this cruise becomes a blur") leans into loss-aversion and nostalgia rather than neutral, informational phrasing, betting that reminding someone what they stand to lose (the memory) motivates action more than a plain reminder to log in.
Across 25 sends, this email saw a 72% open rate and a 28% click rate - notably outperforming the Mid-Cruise email's 47% open rate and 12% click rate (n=43), which used more neutral, informational phrasing ("Update your cruise log"). That gap is a real signal that the loss-aversion/nostalgia framing is doing real work here, not just filling space - though with sample sizes this size, it's a promising early result worth continuing to test as volume grows, not a settled conclusion.
See the Mid-Cruise companion piece for the full comparison.

Feature eNewsletter:
Re-Engagement Through Value Discovery

View Email Here
This email served a dual purpose in the customer lifecycle: reactivating dormant accounts and driving feature adoption. Rather than a simple "here's what's new" announcement, the strategic goal was reminding users of an account relationship they may have forgotten about, then giving them a concrete reason to re-enter the product, new functionality they hadn't yet experienced. The underlying bet was that feature announcements are more effective as a re-engagement lever than as a pure product-update broadcast: users who left an account dormant need a reason to come back, and "something new exists that you haven't tried" is a lower-friction hook than a generic win-back message. This ties feature communication directly to activation and retention strategy, not just product marketing.
Across 293 sends (July 7–10, 2026), this email saw a 39% open rate and an 8% click rate. That's a meaningfully larger sample than the cruise touchpoint pair, which makes it a useful baseline for comparison but, on its own it points to an opportunity rather than a win: with roughly 6 in 10 recipients never opening, the subject line's anticipation angle isn't yet doing enough to break through. Another factor could be that the timing of the email not being tied to an activity has weaker performance.

The Fresh Basket:
Personalized Re-Order Campaign (Concept Piece)

View Email Here
I built this concept email to demonstrate the full arc from marketing rationale to execution not just design, but the thinking behind each decision. The strategic premise: recurring-order businesses succeed by reducing the friction and mental effort of reordering, not by re-selling the value proposition every time. The headline leads with a direct, low-effort question a nudge rather than a pitch because a returning customer doesn't need to be convinced. The body copy stays brief and reinforcing rather than persuasive, since the goal is recall, not conversion from scratch. The centerpiece is a three-column recommendation module built around purchase history, using mobile stacking (responsive columns that collapse to a single full-width column on smaller screens) so the recommendations stay scannable on any device — treating personalization as a conversion lever rather than a decorative feature, since showing customers exactly what they'd likely want reduces the number of decisions between "I got this email" and "I placed an order." I coded the email myself to a production-ready standard, so the piece demonstrates both the marketing logic and the technical fluency to execute it without a handoff.

Wine&Paint Not🖌 Night (Concept Piece)

View Email Here
This concept email was built to demonstrate persuasive design layered on top of technical execution by using an emotional hook and a structured layout to guide returning customers toward a higher-value event. The strategic premise leans on pathos: for a paint-night studio, the product isn't really the painting, it's the shared experience of good company, creativity, and low stakes.
This piece takes that same emotional territory and raises the stakes slightly by introducing a coveted expert instructor, reframing a familiar recurring activity ("another paint night") into something more exclusive ("a master class with someone sought-after").
The structure mirrors how that decision actually gets made: lead with the core ask (a prominent sign-up button after covering who/what/where/when, since returning customers don't need to be sold on the studio itself), then follow with credibility-building content an artist portrait and bio for anyone who needs a bit more conviction before committing. Giving that second section its own two CTAs (learn more about the artist, or go straight to the event page) accounts for two different reader mindsets without forcing everyone through the same path.
Technically, the piece uses mobile stacking on its two-column module with two independently-functioning buttons - responsive columns that collapse to full-width on smaller screens - built to hold up across email clients and demonstrating the coding sophistication behind what reads as a simple layout.

Horticultural Home 🌿
Seasonal Newsletter/Promotion Hybrid (Concept Piece)

View Email Here

This concept email blends editorial content with a seasonal sales push, using a pop-culture hook ("Winter Is Coming") to land at the exact moment the audience's problem becomes urgent the seasonal drop in light that stresses light-sensitive houseplants.
Rather than opening with a product pitch, it leads with an educational hook: a chart showing which plant species tolerate low light and which will suffer without supplemental light. This positions the brand as a knowledgeable resource before it positions itself as a seller, which builds trust with a plant-owner audience such as myself.
The species chart also lets the email serve two audiences at once, people whose plants can survive as-is (pointed toward "adopt low-light friendly plants") and people whose plants need help (pointed toward "upgrade your light setup") - segmenting intent within a single send rather than sending two campaigns.
The closing module reframes the purchase decision around the reader's actual motivation (aesthetics, plant health, or budget), acknowledging that "buy a grow light" isn't a single use case.
From a production standpoint, the piece showcases some of my plants and a data-driven comparison chart. I wanted to lean into my creativity and technical ability to code a structured data display in email which is a non standard layout for most marketers.

Take-a-Hike!:
Serialized Destination Newsletter (Concept Piece)

View Email Here
This concept newsletter is built around a serialized content strategy - treating each national park as one installment in an ongoing subscriber relationship, rather than a one-off send. The Grand Canyon edition is structured to serve two reader states at once: hikers ready to commit to a specific trail right now, and hikers who just want to stay engaged with the community and content.
Each trail module (Bright Angel, South Kaibab) pairs a strong visual with practical trip-planning detail (difficulty, mileage context, what makes it worth doing) then routes directly to the authoritative source (NPS trail guides) rather than a generic "learn more," which builds trust by not gatekeeping information the reader actually needs.
The closing module does the retention work: teasing next week's Zion installment with a hook (what gear, when to go, kid-friendly options) that pre-sells the next email before this one is even closed, while also giving readers a path to back-catalog content if they want more now.
Technically, the piece uses mobile stacking — responsive two-column modules that collapse to a single full-width column on smaller screens — a layout decision that matters here specifically because trip-planning content is often read on a phone, mid-trail-research, not at a desk. The photography is 100% original, reinforcing brand authenticity for a niche, enthusiast audience that would clock stock imagery immediately.

Portfolio Email

View Email Here
Honestly, this one was a fun side quest. Just me following my curiosity and playing around. I figured - I make emails for a living, and I have a portfolio site, so why not make the portfolio itself an email? The photo is me playing holochess on a recreation of the Millennium Falcon - another slice of what I'm into outside of work, same as the paint nights, houseplants, hiking, and healthy food running through the rest of this portfolio.
On the technical side, I used this one to break out of the box - literally. Most emails live inside a safe little 600–640px container, but I built a full-bleed dark purple background section that stretches edge to edge, which takes extra coding know-how to get rendering consistently across email clients.

Mid-Cruise Nudge: In-the-Moment Data Capture

View Email Here
This is the first half of the two-part cruise touchpoint strategy — the companion to the Post-Cruise Nudge, sent while the guest is still actively on their trip. The strategic aim here is different from its post-cruise counterpart: this one isn't about emotional re-engagement, it's about data capture at the moment details are most accurate. The copy leans on immediacy ("log it while it's fresh," "the details are sharpest right now") because memory decay makes good data a challenge — the longer you wait, the vaguer and less useful the input gets. The tagging prompt in particular is designed with a second-order use case in mind: those tags become structured data that could eventually power cruise recommendations for other users based on real traveler-generated input.
Across 43 sends, this email saw a 47% open rate and a 12% click rate. Lower than our Post-Cruise email's 72% open rate and 28% click rate (n=25).
We suspect that the neutral, informational framing here ("Update your {dynamically generated cruise} cruise log") appears to underperform the loss-aversion/nostalgia framing used in Post-Cruise ("Before this cruise becomes a blur"). However another factor could be the timing of the send. Many users will be using mobile devices on a ship when they recieve this email. That means subject lines are likely to get cut off. On many ships cruisers need to pay for internet packages and so many cruisers may not have internet access on their voyage.
A step I would like to work on is setting up an A/B test to see if the long subject line, the tone or the personalization of the subject line impact numbers.
Our current challenge is low volume sends to get quality data.
Therfore I am looking to change one variable the tone and compare, that data against current data.

Unicoder

View Site Here
This tool started as a solution to a real, recurring operational problem I saw across two different employers: email developers were manually cleaning up text, fixing casing, converting special characters and Unicode to email-safe HTML, stripping stray line returns, adding line breaks, superscripting legal characters, unlinking auto-linked addresses and phone numbers — by hand, every single build. That's exactly the kind of repetitive, error-prone work that doesn't need a person doing it manually every time. I rebuilt the tool as a standalone, broader-concept version (not tied to any one employer's specific templates or systems) using regex-based find-and-replace logic to automate each of those cleanup steps into a single interface. At minimum, it saved me a couple of minutes on every build, and at least one coworker picked it up and used it regularly too. It's a small example of a bigger instinct I have: when I notice a manual process eating time across a team, I'd rather build the fix than keep tolerating it.

Splash or Cash:
Campaign Concept & Internal Ideation Win

This was a campaign concept developed during an internal ideation competition at my first marketing role, open to a marketing department of about 12 people. Working on a team of three, we developed "Splash or Cash" — a promotional concept where a buyer purchasing a deal could choose between a pool-related perk or cash to put toward their home, giving prospects a choice between two very different but equally motivating incentives rather than a single one-size-fits-all offer. I came up with the slogan myself. For each concept, I took our graphic designer's print design and reinterpreted it into our email templates, translating a static, print-oriented layout into something that would actually hold up in an inbox. I also built a website to present our campaign ideas as part of the pitch — packaging the strategic thinking (the offer structure, the messaging, the target appeal) into something reviewable rather than just a verbal pitch. Our team's concept won the internal competition. I'm not certain the campaign ultimately ran as a live send, but the win itself is early evidence I could generate a marketing concept that held up against internal competition — even this early in my career.

Tic Tac Toe (College Project)

View Site Here
Originally built in college as a JavaScript learning exercise. The focus being on creating a playable Tic Tac Toe game that offered either human vs human player or human vs. computer player games, with added features such as match modes and score tracking. The computer opponent ran on random-placement logic.
Coming back to it, I used Claude Code to rapidly rebuild the computer opponent's logic, adding a true minimax algorithm with adjustable difficulty (Easy vs. Hard/Unbeatable), rather than the original random moves.
What mattered most to me about this update wasn't just fixing the gap - it's what having AI-assisted development freed me up to focus on.
Instead of spending my limited time wrestling with the algorithm itself, I could put that energy into the user experience: difficulty selection, cleaner score tracking, and match-mode options that make the game more fun to play, not just technically functional. I was able to add in responsive design so a user could have a good experience on mobile. Along with adding more UI elements that make website interaction more clear.
It's a small example of how I think AI tooling changes the job, less time on raw implementation, more time on the design and experience decisions that actually matter to the end user.

Zap-A-Ghost (2015/2016)

View Site Here
Originally built during a weekend JavaScript class at a community college, early in my career, as a way to enhance coding skills outside of email work. The class project was a standard whack-a-mole game, but even in a coding-focused class, my background in Interactive Media Studies pulled me toward making something more expressive than the default assets - a custom ghost icon I designed in Adobe Illustrator, a Halloween theme, and a "Ghost in the Machine" title screen with a red alert-style light-up effect.
Coming back to it years later, I used Claude Code to fix what the original version was missing: it wasn't responsive, and there was no difficulty curve, just one fixed pace for everyone. I added Easy/Medium/Hard difficulty tiers that adjust how fast the ghosts appear and how long they stay on screen, and made the layout responsive so it actually works on mobile. Same pattern as my Tic Tac Toe refresh - using AI-assisted development to move quickly through the technical fixes freed me up to focus on making the game feel more complete, not just functional.

Advanced Medical Laser Center:
End-to-End Client Website Build

View Site Here
This was a full client engagement from the ground up not just a build, but the discovery and design thinking that came before any code was written. Working directly with the client, I made the brand-level decisions (color palette, page layout, content and messaging) needed to represent a small, locally-owned medical aesthetics business online, then backed that up with original photography of the client's own equipment rather than generic stock imagery, which matters for a local service business where trust and specificity sell better than polish alone. On the technical side, I hand-coded a custom accordion component for the FAQ section a deliberate UX choice to keep a content-heavy page (treatment info, policies, pricing questions) scannable rather than overwhelming a visitor on a small business site where most traffic is likely on mobile.
I also configured a domain-based business email for the client as part of the engagement, and iterated on the site over multiple rounds of client feedback, managing the relationship and the revisions independently, without a project manager or account lead in between.

Hyperdrive Agile:
Ongoing Site Maintenance &Client Management

This was an ongoing engagement rather than a one-time build. I maintained an event site for a client hosting Agile-focused conferences. The work centered on keeping the site current as each event cycle changed: updating speaker images and event dates, and designing a logo banner to refresh the site's visual identity. I took content and design feedback directly from the client and translated it into updates myself, serving as the sole point of contact for a period of time — meaning the client had one person to bring feedback to rather than routing through a larger team.
This kind of engagement demonstrates reliability and responsiveness over time, working within an existing site rather than starting from scratch.

AgileCamp: Ongoing Site Maintenance & Client Management

A parallel engagement to Hyperdrive Agile, maintaining a second event site for the same client's Agile-focused event portfolio. Work included updating speaker images and event dates as AgileCamp's own event cycle changed, along with logo banner design to keep its visual identity current and distinct from the client's other site. As with Hyperdrive Agile, I worked directly from client content and design feedback and served as the sole point of contact managing the relationship for a period of time - running two related but separately-branded sites for the same client simultaneously.

Kristen Gilles

Marketing technologist

Hi, my name is Kristen: