I build the systems that turn first‑time buyers into regulars.

Churchill's needed a clear plan for growing repeat business: who its best customers were, how to keep them coming back, and what on the website was getting in the way of a purchase.

I used GA4 site analytics, a competitive audit and customer segmentation to set growth benchmarks, and the store's POS data to see who was buying and how often. From there I built a phased roadmap and worked through it directly with ownership.

A full customer lifecycle framework in Mailchimp: a 3-part welcome flow for new customers and a 3-part win-back flow for lapsed ones. I handed both to ownership. The building was sold before rollout, so they were never launched.

A tiered loyalty program designed from the POS data, with its activation plan built in HubSpot CRM.

On the website: UX recommendations for navigation, A/B test proposals, HTML/CSS fixes to reduce friction, and a plan for an online store with age verification.

Sweet Boba House needed marketing that would build awareness with nearby students and young professionals, and bring first-time customers back.

Our team built a strategic marketing plan covering product, pricing, distribution and promotion. My job was to turn it into something measurable: the goals framework and a phased roadmap with clear targets.

The plan was the deliverable. After the engagement ended, I kept going on my own with a paid media test and a lifecycle email sequence, starting with a seasonal email built around the plan's winter flavors.

Paid media strategy Concept, not executed

A small test aimed at two goals the owner had already agreed to: TikTok engagement and repeat purchases. A local boba shop doesn't need broad awareness ads. It needs nearby people to come in once, then come back.

I left Google Search out on purpose. Boba demand is impulse-driven, and few people search for it with purchase intent. Success is measured by cost per loyalty signup, cost per redeemed offer (the “show this post” mechanic makes redemptions countable at the register), and saves, shares and comments as early signals.

United Way of Greater Chippewa Valley wanted to understand what actually drives college students to volunteer with its affiliated programs, so it could reach them more effectively.

I co-led a four-person team that designed and ran a survey of 63 UW–Eau Claire students on their motivations and the causes they care about most.

We grouped the open-ended answers into seven categories and averaged the ratings for ten causes. Two findings stood out.

Students were moved most by connection and by how volunteering made them feel, and the causes they rated highest were youth support and animal welfare. We presented three recommendations to program stakeholders, each tied to a specific finding.

The pouch category was growing fast, and the major brands I audited (Zyn, On!, Rogue) all competed on flavor variety, strength tiers or lifestyle. None of them led with the product itself: how it feels in the mouth and what goes into it.

I audited how each brand positions itself. All three were fighting over the same ground: more flavors, more strength options, a more discreet lifestyle image. That's a crowded, red-ocean contest where the only way to win is to out-spend or out-flavor the leader.

That left the mouth itself unclaimed. I treated it as a blue-ocean opening, a position no one in the set was claiming, so XYLO wouldn't have to beat Zyn at its own game. It would compete on a different question: how the pouch feels and what's in it.

The position is a focus, not a health promise. XYLO is “the mouth-first pouch”: it puts how the product feels and what's in it at the center of the brand, and its only product claims are facts: made with xylitol, no sugar, one 3 mg strength. It never says it's better for your mouth, teeth or gums. On a nicotine product, any suggestion of lower risk or a health benefit is a modified-risk claim that needs FDA authorization first. None of the brands I audited put the mouth at the center at all. I built it out as a full product line, with White XYLO as the hero product:

I took XYLO through a full spec ad sprint: a product truth sheet, five ad angles, and seven finished assets. Two reels, three alternate hero openings, a carousel and a static, all cut to a beat with captions and sound.

One call shaped all of it. On a nicotine product, “For Mouth Health” reads as a modified-risk claim, which needs FDA authorization. So the positioning got louder (“the mouth-first pouch,” the smile) while the claims stayed provable (“made with xylitol, no sugar”), with the required nicotine warning on every ad.

Delivering for Domino's, I kept running into the same small problem, and it had nothing to do with driving. It happens when the order comes out of the insulated hot bag. Smaller products ride on top of the stacked boxes, and on the way out they catch on the top of the bag and slide off. Either the customer and I end up doing a second hand-off to collect what fell, or they land on the ground.

It isn't the worst problem a delivery can have. But it's the one moment the customer actually meets the brand, and it makes Domino's look unprofessional and the driver look clumsy, when the driver did nothing wrong.

The easy answer is “be more careful pulling the order out,” but that blames the person holding the bag for a packaging gap. Nothing on a stacked order holds the top items in place when the stack moves, and driver technique is the hardest part of delivery to standardize. The box is the one thing the company fully controls before an order leaves the store, and it already has an air hole in the back panel doing nothing but venting.

The Slide Stopper is a small cardboard backboard that folds once and slides into that existing air hole. Folded into place, it forms a raised lip along the back of the box, so anything riding on top, like sides, drinks or a smaller box, has something to stop against instead of sliding off as the stack comes out of the bag. It needs no adhesive, no new tooling and no box redesign. After the Photoshop renders, I hand-cut a cardboard version and fit it into real Domino's boxes to check the fold and the notches actually seat the way the render assumed.

The prototype fits the box the way the render predicted: the notches seat in the existing air hole, and the fold holds its shape under a stacked box. What it hasn't had yet is the test that matters, pulling a loaded stack out of an actual insulated hot bag with smaller products on top, over and over, and counting what stays put. That's the next step, not a finished result.

If it holds up, the payoff is small but real: no dropped products, no second hand-off, and a driver who looks as capable at the door as they are. It's a cheap fix for a moment that happens on the customer's doorstep.

Most Valentine's-themed sneaker drops bolt on a generic heart, clip art dropped next to the Swoosh instead of grown out of it. I wanted a heart that came from the brand's own mark instead of sitting beside it, so the design would still read as Nike at a glance instead of a seasonal sticker layered on top of one.

The Swoosh already has the right kind of curve. I mirrored a second Swoosh vertically and set it over the first, so the two curves met at the tip and the tail and closed into a heart outline, built entirely out of Nike's own mark rather than an unrelated shape. From there it came down to stitching: a dense satin-stitch fill for the OG edition, or a clean, open running-stitch outline for Half Hearted.

The same trick works a second time. Flip the heart horizontally against a mirror of itself and the two shapes cross in the middle, closing into a figure eight, an infinity symbol built from the same curves as the OG heart. Once that two-heart unit existed, I rotated it around a center point. Four copies, set 45° apart, give eight hearts that close into an eight-petal flower: the same heart shape repeated and rotated instead of drawn from scratch.

OG, Half Hearted, Infinity and Flowers are the same underlying shape at four stages of the same idea, not four unrelated designs. Once the flower pattern existed, running it across a full range of colorways instead of a single hero color was straightforward, the kind of system a real product line needs rather than a one-off graphic.