After this lesson you should be able to
- Calculate the replacement rate your profile needs before setting a growth target
- Describe what a standing link program contains beyond outreach
- Staff and budget a program rather than a series of projects
- Judge program health by net referring domains rather than gross links
The project trap
Link building is usually bought as a project. Someone approves a quarter, a campaign runs, links arrive, everyone is pleased, and the budget moves on to the next thing. Six months later the rankings have drifted back and the conclusion drawn is that the links did not work.
The links worked. They stopped. That is a different failure and it has a different fix.
The project framing assumes links are an asset you buy once, like a piece of equipment. They behave much more like a subscriber base: acquired continuously, lost continuously, and healthy only when acquisition exceeds loss. Once you see the profile that way, the question changes from how many links do we need to how many do we need per year to stay level, and then how many above that to grow. Almost nobody runs that second calculation, and it is the one that determines whether a program is adequately funded.
The replacement rate argument
Here is the arithmetic, using data from a real profile rather than a rule of thumb. Across 1,301,839 links from 22,260 referring domains on my own site, measured over twenty years:
- The median referring domain stops linking after 1,080 days — just under three years.
- 49.6% of all links ever recorded are lost. Not broken, not redirected: gone.
- Survival scales with quality. A Trust Flow 0 domain has a median life of 859 days; a domain at Trust Flow 61 or above lasts a median of 3,353 days, roughly four times as long.
A median life of about three years means that, in steady state, a profile sheds roughly a third of its referring domains every three years, which is close to one sixth of the profile per year once you account for the fact that newer links are more fragile than older survivors. To stand still, a site with 300 referring domains needs to acquire something like 50 new ones a year. To grow at all, it needs more than that.
Read that against the previous lesson's expectation of 5 to 15 genuine referring domains a quarter and the implication is uncomfortable but clear: for many sites, a well-run quarterly campaign is roughly a replacement operation. Growth requires either a higher rate or better links, and the survival data says better links are the cheaper lever, since a high Trust Flow domain lasts about four times as long as a Trust Flow 0 one and therefore needs replacing a quarter as often.
That is the strongest argument I know for quality over volume, and it does not rely on any claim about ranking weight at all. It is purely about how long the link stays there.
Two practical consequences follow. First, calculate the replacement rate before you set a growth target, because a target set without it is arithmetic done in the wrong order. Second, when you compare two acquisition options of similar cost, weight them by expected life rather than by count: forty fragile links and ten durable ones are not the same purchase spread over three years, and the second is usually the better one.
What a standing program contains
A program is not a campaign that never ends. It is a small set of recurring activities with different cadences, most of which are not outreach.
- Continuous reclamation. Comparing referring-domain snapshots on a schedule, finding what has quietly disappeared, and asking for it back. This is the highest-yield recurring work available and module 8 explains why.
- Mention monitoring. Catching unlinked mentions while they are fresh, when the writer still remembers the piece and the request costs them thirty seconds.
- One or two active acquisition tactics, run on a rolling basis rather than in bursts, with a standing prospect list that is topped up rather than rebuilt.
- A publishing commitment — whatever cadence you can sustain for material that earns links without being asked. Once a quarter is enough if it is genuinely good.
- Relationship maintenance. Staying in contact with the people who have linked before, because the second link from an existing contact costs a fraction of the first from a stranger.
- A quarterly review of net referring domains, not gross links.
The proportions shift as a site matures. A new site spends most of its effort on acquisition. An established site with a large profile can spend half its effort on reclamation and maintenance and still come out ahead, because it has more to lose than it has room to gain.
Staffing and budgeting a program
Programs fail for budget reasons more often than technical ones, and the pattern is consistent: the work is funded in bursts, the person doing it is reassigned between bursts, and the relationships built in the first burst are cold by the second.
Three structural principles help. Continuity beats intensity. One person spending a steady day a week for a year will out-produce four people spending a month, because outreach compounds through relationships and relationships need continuity. Name an owner. Diffuse ownership means reclamation never happens; it is the first thing dropped when everyone is busy, and it is the highest-yield thing on the list. Fund the boring half. Monitoring, snapshots and reclamation are unglamorous and produce no story to present, which is exactly why they get cut, and cutting them is what turns a program back into a series of campaigns.
On cost, I will describe the shape rather than figures, which go stale. The dominant cost in a serious program is skilled human time, and it splits roughly between research, creation and correspondence. Tooling is a small fraction. Anything that promises to replace the correspondence with automation is proposing to remove the part that produces the links.
How to tell the program is working
The measure is net referring domains over time: domains gained minus domains lost, tracked quarterly, compared against the replacement rate you calculated at the start. Everything else is diagnostic detail.
Three patterns and what they mean:
- Net positive and rising, with the gains skewed toward higher Trust Flow. The program is working and the profile is getting more durable at the same time. This is the target state.
- Gross acquisition healthy but net flat. You are replacing, not growing. Either the acquisition rate needs to rise or the links being acquired are too fragile. Check the Trust Flow distribution of new domains against the losses.
- Net negative despite activity. Loss is outrunning acquisition, which for an older profile with a lot of legacy links is common and is not necessarily an emergency. Prioritize reclamation over acquisition for a quarter and re-measure; recovered links are cheaper than new ones.
One caution about vanity. Gross link counts will mislead you at this level more than anywhere else, because a single sitewide placement can add tens of thousands of links without adding a single relationship. On my profile, one domain contributed 16.8% of all links from what was, editorially, one person's decision. Count domains, judge durability, and let the link count be a footnote.
Questions
Do links really expire?
Not on a timer, but they disappear at a predictable rate. On a twenty-year profile of 1,301,839 links, 49.6% are gone and the median referring domain stopped linking after 1,080 days. Pages get redesigned, sections get retired, content management systems get replaced, and links vanish as a side effect. The rate is remarkably steady even though each individual loss is unpredictable.
How many links do I need per year just to stand still?
Roughly a sixth of your referring domain count, as a working estimate derived from a median domain life near three years. A site with 300 referring domains needs about 50 new ones a year to hold position. Higher-quality links reduce that burden substantially, since domains at Trust Flow 61 and above last around four times as long as Trust Flow 0 domains.
Is it cheaper to keep links or to get new ones?
Keeping them, by a wide margin. A reclamation request goes to someone who already linked to you once, references a specific page they already published, and asks for a small correction. A new link requires finding a stranger, giving them a reason to care, and persuading them. Reclamation is the least glamorous and most profitable work in a standing program.
When should a campaign become a program?
As soon as the first campaign produces links worth keeping. The switch is administrative rather than technical: a named owner, a recurring calendar entry, a standing prospect list, and a quarterly snapshot comparison. If you wait until the profile is large enough to feel worth maintaining, you will already have lost a year of decay you could have prevented.