Question 1

Our numbers disagree across platforms — can you fix that?

Accepted Answer

Usually yes. Most discrepancies come from event architecture and attribution windows rather than the data being wrong. We audit the measurement layer first, then rebuild GA4 and GTM tracking to a single agreed definition of a conversion.

Question 2

Can you guarantee coverage?

Accepted Answer

No, and anyone who does is either paying for it or misleading you. What we can commit to is a story worth covering, the right journalists approached properly, and the responsiveness that makes a publication come back to you next time. Earned media is earned.

Question 3

What happens if something goes wrong publicly?

Accepted Answer

We work to a prepared position rather than improvising. That means agreed spokespeople, holding statements drafted before they are needed, a clear escalation path and a decision on what you will and will not say. Speed matters in a crisis, and speed comes from preparation, not from talent.

Question 4

Is PR still worth it when everything is digital?

Accepted Answer

More than it was. Independent third-party coverage is one of the strongest trust signals both to people and to AI answer engines assessing whether a brand is credible. A well-placed article outlives a campaign burst and keeps working in search for years.

Question 5

Do you buy traditional media as well as digital?

Accepted Answer

Yes — television, radio, press and out-of-home alongside digital and programmatic, planned and bought by the same team so the mix is optimised together.

Question 6

How often should we be posting?

Accepted Answer

Cadence matters less than consistency and fit. A brand posting three considered pieces a week that suit the platform will beat one posting daily filler. We set a rhythm your team or ours can genuinely sustain, because an ecosystem that stops for a month costs more reach than it ever gained.

Question 7

Do influencers actually work for B2B?

Accepted Answer

Yes, but not in the consumer sense. In B2B the equivalent is credible practitioners with a specialist audience — a smaller, expert following converts far better than reach. We look at audience composition and demonstrated authority, never follower count alone.

Question 8

Who owns the content you produce?

Accepted Answer

You do. All assets, source files and accounts remain your property, and we hand over working files as a matter of course. If you ever move on from us you should leave with everything, not a locked account and a folder of exports.

Question 9

How do you measure social when it does not drive direct sales?

Accepted Answer

By its actual job. For most brands social builds recognition and consideration, so we track engaged reach, audience growth quality, and assisted conversions through GA4 — then look at whether branded search and direct traffic rise alongside it. Attributing a dessert purchase to a single post is fiction; showing that demand moved is not.

Question 10

What budget do we need to start?

Accepted Answer

Performance media becomes measurable from roughly $3,000–5,000 per month in media spend; below that, learning periods take too long to optimise reliably. We will say so plainly if paid media is not the right first investment.

Question 11

How do you report performance?

Accepted Answer

GA4 and platform data consolidated into one dashboard against blended ROI and lifetime value — not last-click.

Question 12

What is Generative Engine Optimization (GEO)?

Accepted Answer

GEO is the practice of making a brand visible and citable inside AI answer engines — ChatGPT, Perplexity, Gemini and Google AI Overviews — rather than only ranking in classic search results. It combines entity clarity, structured data, and content written to be quoted.

Question 13

How is GEO different from SEO?

Accepted Answer

SEO competes for a position in a list of links. GEO competes to be the source an AI model cites in its answer. The disciplines overlap in technical health and content quality, but GEO adds entity definition, schema depth and answer-shaped content.

Question 14

Can you measure it?

Accepted Answer

Yes — we track brand mention rate and citation share of voice across ChatGPT, Perplexity, Gemini and Google AI Overviews, alongside GA4 and Search Console, and report them together.

Question 15

If AI answers reduce clicks, is the traffic worth less?

Accepted Answer

The volume falls but the quality rises. Visitors who arrive after reading an AI answer have already been pre-qualified, and AI-referred traffic has been measured converting at several times the rate of traditional organic. The risk is not lower conversion — it is being absent from the answer entirely.

Question 16

How long before we see movement?

Accepted Answer

Entity and schema work can change how engines describe you within weeks. Citation share against established competitors is a two-to-three quarter programme, because it depends on earned sources as well as your own content.

Question 17

Can you work to an election or launch deadline?

Accepted Answer

Yes. We run compressed campaign windows with rapid-response creative and electorate- or region-level targeting — the model behind our political and advocacy work.

Question 18

How long does a brand project take?

Accepted Answer

Positioning and identity typically run eight to sixteen weeks depending on research depth and how many stakeholders need to align. The variable is rarely design time — it is the time to reach a genuine internal decision about what the business stands for.

Question 19

Do we need a full rebrand or just a refresh?

Accepted Answer

Usually a refresh. A full rebrand is warranted when the business has genuinely changed, when the name no longer fits, or when equity is negative. If your customers still recognise and trust you, discarding that recognition is expensive. We will tell you if a rebrand is not justified.

Question 20

What do we actually receive?

Accepted Answer

Positioning, messaging framework, identity system, and guidelines your team and your suppliers can apply without us in the room. Files are handed over in working formats. A brand system that only we can operate is a dependency, not a deliverable.

Question 21

What is AWS Security Hub and why do you standardise on it?

Accepted Answer

AWS Security Hub CSPM gives a comprehensive view of your security state in AWS and assesses the environment against industry standards and best practices. It collects security data across accounts, AWS services and supported third-party products, correlates findings so the important ones surface first, and presents them in one place. Rather than managing findings from many sources in different formats, everything arrives in the AWS Security Finding Format. Full reference: https://docs.aws.amazon.com/securityhub/latest/userguide/what-is-securityhub.html

Question 22

Which compliance standards can you evidence against?

Accepted Answer

Security Hub CSPM supports multiple standards, including the AWS Foundational Security Best Practices standard developed by AWS, and external frameworks such as CIS, PCI DSS and NIST. Each standard contains controls representing a security best practice, and continuous checks generate control findings you can hand to an auditor.

Question 23

Do we need anything else turned on first?

Accepted Answer

Yes. Security Hub CSPM uses service-linked rules from AWS Config to run most control checks, so AWS Config must be enabled and recording resources. It is also worth knowing that Security Hub only detects findings generated after it is enabled — it does not backfill — and it processes findings in the Region where it was enabled, so full CIS AWS Foundations Benchmark coverage requires enabling it in all supported Regions.

Question 24

Can findings be actioned automatically?

Accepted Answer

Selectively. Automation rules can modify or suppress findings against criteria you define, and the EventBridge integration can trigger custom actions — raising a ticket, or invoking an automated remediation. We automate the safe, repetitive triage and deliberately keep a human approving anything that changes access, alters data or affects a customer account.

Question 25

Is this a one-off audit or an ongoing service?

Accepted Answer

Ongoing. A point-in-time audit tells you the posture on one afternoon. Continuous checks, a tracked security score and consolidated findings tell you the posture today, which is the only version that matters when something goes wrong.

Question 26

Do you work alongside our existing security team or provider?

Accepted Answer

Yes. We commonly own the AWS posture and detection layer while an internal team or MSSP owns endpoints and corporate IT. Findings route to whichever system you already run.

Question 27

Why does an AI project need evals?

Accepted Answer

Because a model update or prompt change can silently degrade quality. Automated evaluations score every change against a fixed set of real cases, which is the single biggest predictor of whether an agent survives in production — reported rollback rates drop from about 47% without eval coverage to roughly 9% with it.

Question 28

Who decides what the AI is allowed to do?

Accepted Answer

You do. We facilitate a scoping session with your risk, legal and operational owners, write the policy down, then implement it as enforceable guardrails rather than instructions in a prompt.

Question 29

Does this apply to a small deployment?

Accepted Answer

Proportionately, yes. A single customer-facing assistant still needs a defined scope, refusal behaviour, logging and a way to turn it off. We scale the control layer to the risk.

Question 30

Is our data used to train public AI models?

Accepted Answer

No. We build on enterprise agreements and private deployments so your content is used to answer your questions, not to train public models. Access control and audit logging are part of the design.

Question 31

What if AI is not the right answer?

Accepted Answer

We start with a feasibility check on your real data and recommend conventional automation when it will win. We would rather lose the AI scope than ship something that erodes trust.

Question 32

What does "agentic" actually mean here?

Accepted Answer

An agent carries out a multi-step task — read a brief, pull records from your systems via MCP, draft an output, flag exceptions for a human — instead of answering a single prompt.

Question 33

How do you know a design will work before you build it?

Accepted Answer

We test it. Research with real users, prototypes people can click, and behavioural evidence rather than internal opinion. It is far cheaper to discover a flow is confusing in a prototype than after it is engineered.

Question 34

Is accessibility included or extra?

Accepted Answer

Included. Contrast, keyboard navigation, focus states, semantic markup and screen-reader behaviour are part of the work, not a compliance pass at the end. It is also increasingly a legal expectation, and retrofitting it costs several times what building it in does.

Question 35

What if our team disagrees about the design?

Accepted Answer

That is normal and usually productive, but it needs evidence rather than seniority to resolve. We use research findings and testing to settle disagreements about what users need, and reserve taste debates for the things that genuinely are taste.

Question 36

Is this just SEO with a new name?

Accepted Answer

No. Classic SEO competes for a position in a list of links. AI search decides whether your brand is cited in an answer where no list appears. That requires entity clarity, structured data and content written to be quoted — related to SEO but not the same discipline.

Question 37

How do we know if AI engines are citing us?

Accepted Answer

We track it directly, prompting the major answer engines on the questions your buyers actually ask and recording whether you appear, how you are described and who is cited instead. It is a genuine measurement problem, not a guess, and the description matters as much as the mention.

Question 38

What if the AI describes our business incorrectly?

Accepted Answer

That is the most common and most damaging failure, and it is usually caused by weak entity signals — inconsistent naming, missing structured data or thin authoritative content about what you actually do. Fixing the source material is what corrects the description; there is no direct edit button.

Question 39

Will this break our existing systems?

Accepted Answer

Not if it is done properly. We map the current state first, build against a staging environment, and use idempotent operations with retry and dead-letter handling so a failed message is never silently lost. Nothing goes near production until the failure paths have been tested, not just the happy path.

Question 40

What if a vendor has no API?

Accepted Answer

There is almost always a path — scheduled file exchange, database replication, a middleware layer or, occasionally, a well-behaved robotic process. It is less elegant and we will be honest about the trade-offs, but a missing API rarely means the integration cannot happen.

Question 41

Who owns the integration afterwards?

Accepted Answer

You do, and it is documented so your team or another supplier can maintain it. We can support it ongoing if you prefer, but that should be your choice rather than a consequence of nobody else being able to understand it.

Question 42

Will my ERP stay in sync with the storefront?

Accepted Answer

Yes. We integrate Dynamics 365 Business Central, Xero or QuickBooks via APIs and webhooks so inventory, pricing and fulfilment stay current — the pattern behind Wafex reducing order processing time by 50%.

Question 43

Can you handle wholesale and retail pricing together?

Accepted Answer

Yes — multi-tier pricing, customer-specific catalogues and quick-order workflows on Shopify Plus.

Question 44

Why AWS rather than another cloud?

Accepted Answer

We standardise on AWS because our team runs it daily — Cognito for identity, Lambda for serverless logic, S3 for encrypted storage and CloudWatch for observability. Deep familiarity in one cloud produces more reliable, cheaper systems than shallow coverage of three.

Question 45

Can you take over an existing platform?

Accepted Answer

Yes. We start with an architecture and cost review, stabilise what is running, then modernise incrementally rather than forcing a rebuild.

Question 46

How long does an enterprise web build take?

Accepted Answer

Most enterprise platforms run 10–16 weeks from discovery to launch, depending on integration scope. We work in two-week increments with a demoable build at the end of each, so you see progress rather than waiting for a reveal.

Question 47

Who owns the code?

Accepted Answer

You do. We hand over the repository, documentation and infrastructure-as-code, so you are never locked to us to make changes.

Question 48

How do you keep a build secure?

Accepted Answer

We deliver aligned to SOC 2 and ISO 27001 practices: least-privilege access, encrypted data at rest and in transit, dependency scanning in CI, and audit logging from day one.

Everything under one accountable roof.

Build the system that carries the business

Enterprise Web & Application Development

Cloud & SaaS Architecture

Omnichannel Commerce Solutions

Systems Integration

AI Search, MarTech & Data Intelligence

UX, Behavioural Design & Brand Systems

AI Enablement & AI-Native Delivery

AI Governance, Evals & Assurance

Cyber Security Services

Then create the demand it has to carry

Brand & Identity

Campaign Strategy

AI Search & GEO

Paid Media & SEM

Social & Influencer

Media Buying

PR & Communications

AI Analytics & Attribution

No surprises, at any stage

Discovery & scope

Design & architecture

Build in increments

Launch & measure

A deliberately focused stack

Let's scope the right mix for your goals.