<!-- CloudEngage · Adam Parish -->
> Redesigning CloudEngage, a website personalisation product: self-serve sign-up, faster content creation and reusable audiences. Active personalisations grew 145% in three months.

(03) CloudEngage / Web personalisation

# CloudEngage.

Personalisation a small marketing team can run on its own.

CloudEngage lets marketing teams change what their website shows to
each visitor. The product was six years old, built by developers with
no design guidance, and customers were leaving. I led product design:
research with customers, a self-serve sign-up, a faster way to make a
personalisation and choose who sees it, and the product’s first design
system.

Context CloudEngage

Role Design & front-end

Year 2019

Outcome +145% active personalisations

[Open the app](https://atomparish.com/lab/cloudengage/)

01

## The product I inherited

Six years of features, one dashboard of campaign tables, and no
design guidance behind any of it. Interviews and usability tests
showed where it failed: content was slow to make, audience setup
was unclear, and every campaign rebuilt its audience from scratch.

The dashboard when I joined: views, actions and conversions per
campaign

A campaign report by city, over a map of visitors

02

## Sign up without a sales call

Sales set up every new account by hand. The new flow takes a team
from an email address to a live install on its own: account,
details, company, sites, plan and the code snippet.

Design note The install
step has tabs for a custom site, WordPress and Shopify, and a way
to send the snippet to a teammate, because the marketer signing up
is often not the person who can paste it.

A free account, with email, Microsoft or Google

The company: industry and team size

A plan, monthly or yearly, or hand billing to a teammate

The snippet for a custom site, WordPress or Shopify

03

## From a blank page to a live popup

Making content was the slowest part of the old product. Now a
marketer picks a style, then a template, then edits it in place
with their own page behind it.

Design note Templates
filter by goal as well as by layout, such as capturing email or
reducing cart abandonment, so a marketer can start from what the
campaign is for.

Eight styles, from a toaster to a full-screen takeover

Templates filtered by goal and season

Editing the heading in place, with its type settings above

04

## Audiences you build once

Teams told us they rebuilt the same audience for every campaign. A
personalisation now holds who, when and how often on one page, and
an audience can be saved as a segment and reused.

Design note Each targeting
option reads as a plain sentence, such as visitors browsing from a
specific location, instead of a raw rule field.

One personalisation: variants with their results, who sees it
and when

Location and weather added from the targeting options

Saving the audience as a segment to reuse

05

## A design system for a team without a designer

I set up the product’s first design system, from form controls to
buttons, and moved every area of the product onto it, settings and
site install included.

Site settings on the system: install status, analytics and
privacy support

06

## Know who is visiting

The problem under all of it: marketing teams could not tell who
was on their site. The dashboard shows visitors, what they care
about and which pages they read, and the audience list filters
people down to the ones worth targeting.

The dashboard: visitors, conversions, affinities and top pages

Everyone visiting from Spokane, with what each is into

Design system

## CloudEngage

Cool blue-grey grounds with one deep teal-blue brand; square-ish 6
to 10px corners, a violet second series and nothing decorative.

### Typeface

Figtree

Regular 400 · Medium 500 · Semibold 600 · Bold 700

Numbers in Geist Mono

### Type scale
- Display
30 / 36 · 700
- Title
24 / 32 · 700
- Heading
20 / 28 · 600
- Subhead
18 / 28 · 600
- Body strong
16 / 24 · 600
- Body
14 / 20 · 500
- Caption
12 / 16 · 500

### Colour
- Brand
#12739B
Primary action, links, active nav and the one accent
- Brand deep
#0E5F82
Pressed and hover state of the brand
- Brand soft
#E1F0F6
Selected rows, chips, quiet highlights
- Ink
#0E2A3B
Headings and primary text
- Ink 2
#2A4352
Body copy and secondary text
- Ground
#F1F5F8
Page background
- Surface
#FFFFFF
Cards, menus, the editor canvas
- Line
#D5DFE6
Hairlines and card edges
- Positive
#0B7A57
Success, verified install, live
- Danger
#B42318
Errors and destructive actions
- Series 2
#7A5BC7
Second chart series and variant B
- Series 4
#B8620F
Fourth series and caution marks

### Icons · Lucide
- layout-dashboard
- megaphone
- users
- target
- history
- shield-check
- globe
- map-pin
- cloud-rain
- mouse-pointer-click
- bell
- life-buoy

### Components

## A product that had stopped growing

CloudEngage lets marketing teams tailor their website to each
visitor. After six years of features built without a designer, the
product was bloated, retention was low and revenue was drying up.
We set three goals: lift usage through better UX and visual
design, grow the north-star metric of active personalisation
campaigns, and let customers sign up on their own.

## Finding where it broke

I interviewed 17 current customers and 11 marketing managers in
our target market, ran usability tests with 3 existing and 2 new
users, and added a heuristic review, competitive analysis and
product analytics. Our customers were high-performing marketing
teams of three to five people at auto and RV dealerships,
restaurant chains and e-commerce sites, always looking for the
next conversion win and short on time to learn a complex tool.

## Focus over breadth

With a product manager, a tech lead and three engineers, I ranked
the work on a priority matrix and put sign-up, content creation
and targeting first. Two lessons stayed with me: solve one
market’s needs before expanding, and know when to retire a
feature.

## Outcome

In the first month of the new onboarding, 27 of 30 new customers
finished setup without contacting support, and sales no longer set
up accounts by hand. Active personalisations grew 145% in the
three months after the new workflow launched, and we signed a
large auto and RV dealership.

Next project

[ChordConnect →](https://atomparish.com/work/chordconnect/)

Spokane, WA
Back to top ↑
