IDJOY reports that its campaign notes recorded 178 prospects across two South African campaigns: 174 print businesses and four Pretoria dental practices. The notes record 15 sent emails at an unspecified snapshot. The source records are not published, so readers cannot verify the counts independently. They measure research and outreach activity, not replies, meetings, or sales.
About this case study: IDJOY prepared this account from its campaign notes. The prospect records and outreach log are not published, and the snapshot date is not stated, so readers cannot independently check the reported figures.
Updated 4 November 2026: The reported counts are now labelled clearly, the unverified referral-credit promotion and undocumented commission terms have been removed, and the Hyperagent billing and POPIA references have been refreshed.
Lead-generation research identifies suitable businesses, checks public contact details, records a reason for contact, and tracks outreach. IDJOY reports that its campaign notes recorded 178 prospects across two South African campaigns. The counts show research and sending activity, not business outcomes.
IDJOY reports using an AI-assisted, human-reviewed workflow across two South African outreach campaigns. At the unspecified snapshot described in its campaign notes, the lists contained 178 prospects whose email addresses IDJOY says it checked against public business sources. IDJOY reports that it had sent 15 outreach emails by that point.
Those numbers require a clear boundary. A researched prospect is not a sale. It is not a booked meeting. It is not even a sales-qualified lead until the business responds or takes another qualifying action.
This case study describes the research and outreach process. It reports no replies, meetings, revenue or conversion rates.
The campaign snapshot
| Campaign | Prospects in IDJOY’s notes | Emails sent at unspecified snapshot | Main audience |
|---|---|---|---|
| Print and production partners | 174 | 11 | South African print businesses and potential agency partners |
| Pretoria dental practices | 4 | 4 | Pretoria dental and orthodontic practices |
| Total | 178 | 15 | Two focused South African prospect groups |
IDJOY’s notes describe 174 print and production prospects: 11 had received first outreach and 163 remained in the sending queue. The list was divided into three working segments:
- 16 white-label overflow prospects,
- 65 commercial referral prospects,
- 93 standard referral prospects.
IDJOY’s notes describe four Pretoria dental practices, all sent personalised HTML outreach from the company’s business email account. The article does not publish the business names, source links, or capture dates, so readers cannot independently inspect the selection or verify the records.
These figures describe the campaign records at one point in time. They may change as messages are sent, addresses are corrected and prospects respond.
The problem IDJOY wanted to solve
Generic cold outreach wastes the sender’s time and the recipient’s attention.
The team needed a repeatable process that could answer six questions for every prospect:
- Does this business fit the offer?
- Is the contact information public and current?
- Can we identify a real business or website opportunity?
- Can the message refer to evidence rather than a generic compliment?
- Has this prospect already received outreach?
- When should the next action happen?
AI helped with research, classification and drafting. A person remained responsible for verification, the offer, the final message and the campaign record.
The workflow
1. Define the niche and offer before researching
The print campaign did not begin with “find companies”. It began with partnership routes.
Some businesses could refer web projects to IDJOY. Others could use a white-label arrangement: IDJOY would deliver the web work for a partner agency to present under its own brand. The campaign notes describe these routes but do not publish the full referral terms, qualification rules, or payment basis.
The dental campaign used a different offer. IDJOY says it sought practices with strong public feedback and a specific website opportunity. The article does not name the practices or publish evidence for those assessments.
A clear offer changed the research criteria. The researcher could reject businesses that did not fit instead of filling the spreadsheet for the sake of volume.
2. Use geography and business type to control the search
The dental campaign used a Pretoria geographic boundary and targeted dental or orthodontic practices.
The print campaign grouped businesses by partnership fit. This made the list easier to prioritise and allowed the outreach copy to match the commercial relationship.
Narrow criteria also make AI easier to supervise. A request such as “find leads in South Africa” leaves too much room for irrelevant results. A request with an industry, area, minimum evidence standard and required fields produces work that can be checked.
3. Verify the official business presence
Each prospect needed a current official website or a credible public business listing. The workflow used public sources to confirm:
- business name,
- location or service area,
- official website where available,
- public email address,
- phone number where relevant,
- services and market fit.
The process rejected guessed email patterns. It did not create addresses such as info@domain.co.za simply because the domain existed. An address had to appear on an official website or another current public source connected to the business.
That rule reduced the list, but it improved trust in the records.
4. Record a specific reason for contact
Personalisation did not mean adding the prospect’s name to a template.
The campaign record needed an evidence-backed reason. Examples included a website that did not clearly show the full service range, a weak enquiry path, a public profile with strong customer proof that the website did not use well, or a partnership fit based on the company’s production work.
The reason was stored before the message was drafted. This prevented the copy from inventing a problem to justify the offer.
5. Draft from evidence and keep a consistent structure
AI supported first-draft outreach, but it worked from the research record.
The message structure remained consistent:
- identify the business and why it fits,
- name the observed opportunity,
- explain the relevant IDJOY offer,
- give the recipient a low-friction next step,
- avoid invented urgency or exaggerated promises.
The dental messages used personalised HTML. The print campaign used templates matched to the relevant partnership segment. The final messages were sent from hello@idjoy.co.za rather than a free personal address.
6. Update the tracker after every sent message
The spreadsheet acted as the campaign memory. Each completed send could record:
- message ID,
- outreach status,
- first outreach date,
- next action date,
- template version,
- notes or corrections.
This matters because follow-up depends on reliable state. Without it, the team risks contacting the same prospect twice, missing a reply or sending a follow-up too early.
The print campaign planned batches of 10 emails per weekday at 07:55 in the Africa/Johannesburg timezone. A controlled schedule protected quality and made the tracker easier to maintain.
7. Plan follow-up before the first send
IDJOY’s notes describe this planned follow-up rhythm:
- first follow-up after seven days,
- second follow-up after 14 days,
- stop or change the next action when the prospect replies.
A follow-up system should respect opt-outs and avoid repeated contact when the recipient has declined. It should also record replies before another scheduled message runs.
Where AI saved time
AI reduced effort in the parts of the workflow that involve reading, sorting and structuring information.
It helped the team:
- turn research into consistent prospect records,
- classify businesses by campaign segment,
- compare websites against a defined checklist,
- draft outreach from recorded evidence,
- maintain the wording and format across a campaign,
- identify missing fields before a message was approved.
The human review still carried the risk. A person checked the source, business fit, email address, claim and final message.
Where AI could have damaged the campaign
A poorly controlled system could have produced a bigger list in less time and lowered the value of every row.
The main failure modes were:
- guessed email addresses,
- outdated websites or closed businesses,
- generic “your website needs improvement” claims,
- invented personalisation,
- duplicate prospects,
- outreach sent without a recorded status,
- automatic follow-ups after a reply,
- messages that ignored South African privacy and direct-marketing obligations.
The workflow used rejection rules because accuracy mattered more than list size.
What the figures show
IDJOY’s campaign notes describe a structured prospect-research process across two markets. This is a company-authored account, not an independently audited case study.
The company reports that:
- IDJOY recorded 178 prospects across the two lists and says it checked their email addresses against public business sources,
- the lists used defined segments and selection criteria,
- 15 first-outreach messages had been sent at the snapshot,
- each campaign had a tracker and follow-up logic,
- the process combined AI assistance with human verification.
This page reports no reply rate, meeting rate, close rate or return on investment. Those metrics require campaign response data and a defined observation period.
This distinction protects the reader from a common problem in AI lead-generation content: impressive numbers that count generated rows rather than business results.
The measurement plan for the next stage
A useful campaign report should separate activity from outcomes.
Research quality
Track:
- prospects researched,
- contacts verified,
- records rejected,
- duplicate rate,
- percentage with a specific evidence-backed contact reason.
Outreach activity
Track:
- messages sent,
- delivery failures,
- first follow-ups sent,
- second follow-ups sent,
- opt-outs.
Commercial outcomes
Track:
- replies,
- positive replies,
- qualified conversations,
- meetings booked,
- proposals requested,
- projects won,
- revenue and referral fees linked to the campaign.
Only the last group can support a claim that the campaign generated business. The earlier metrics show whether the operating process is healthy enough to reach that point.
Could an AI agent run this workflow?
Not in the historical campaigns described here. IDJOY reports that people reviewed prospect records and outreach before sending. Any future use of an AI agent should keep a human approval step for contact selection, personal-data handling, message claims, and sending.
Hyperagent changed to monthly plans on 26 August 2026. Its current billing guidance says an active monthly plan is required to run agents and use promotional or referral credits. Check the current Hyperagent billing guidance before comparing it with other tools.
When your agency has no time to generate leads
An agency can understand its ideal client and still fail to build a consistent pipeline. Client work takes priority. Research stops. The spreadsheet becomes stale. Follow-ups happen when someone remembers.
IDJOY can operate the research and outreach layer on behalf of an agency. The service can cover:
- campaign and prospect criteria,
- public-source business research,
- email verification without guessed addresses,
- evidence-backed prospect notes,
- segmented outreach drafts,
- approved message sending,
- status and follow-up tracking,
- campaign reporting.
The agency keeps its positioning, pricing and sales relationship. IDJOY handles the repeated work required to keep qualified prospects moving into the outreach process.
A campaign should begin with a small pilot. Agree on the niche, geography, exclusions, required evidence, approval process and success measures. Review the first batch before increasing volume.
Discuss a done-for-you lead-generation pilot with IDJOY
Privacy and outreach controls
Public contact information does not automatically make direct marketing lawful. South Africa’s Protection of Personal Information Act (POPIA) regulates the processing of personal information. Section 69 generally prohibits processing a data subject’s personal information for unsolicited electronic direct marketing unless the person has consented or meets the Act’s existing-customer exception. The Information Regulator’s direct-marketing guidance explains consent, objections, and the distinctions between channels. The rules depend on the recipient and circumstances; get legal advice for a specific campaign.
Operational controls should include:
- use only the data needed for the campaign,
- record the public source,
- protect the spreadsheet and connected accounts,
- honour opt-out requests,
- stop automated follow-ups after a reply,
- review permissions before connecting an AI agent,
- avoid uploading sensitive client or personal information without a lawful and secure process.
AI can perform more actions as tools become more agentic. That increases the need for clear approval boundaries.
The lesson
The reported result is a snapshot of activity: 178 researched prospects and 15 sent emails across two focused South African campaigns. The published account does not establish replies, meetings, sales, or return on investment, and its underlying records are not available for independent review.
For an agency planning a similar campaign, start with a narrow audience, document the source and reason for each contact, review every message, and track responses separately from sends. IDJOY’s pricing page explains how to discuss a lead-generation pilot.